Page 1
i
Evaluation of the Inelastic Rotation Capability of
Flush End-Plate Moment Connections
By
Mark R. Boorse
Thesis submitted to the Faculty of the
Virginia Polytechnic Institute and State University
in partial fulfillment of the requirements for the degree of
MASTER OF SCIENCE
In
Civil Engineering
APPROVED:
T.M. Murray, Chairman
W.S. Easterling T.E. Cousins
March, 1999Blacksburg, Virginia
Page 2
ii
Evaluation of the Inelastic Rotation Capability ofFlush End-Plate Moment Connections
by
Mark Richard Boorse
Committee Chairman: Thomas M. MurrayCivil Engineering
(ABSTRACT)
An experimental investigation was conducted to study the inelastic rotation
capability of flush end-plate moment connections. Seven specimens representing two-
bolt and four-bolt flush end-plate configurations were tested under cyclic loading.
“Quasi-static” or “slow-cyclic” loading histories suggested by SAC and the Applied
Technology Council were used to load the specimens. Experimental results for
maximum moment resisted by the connections were compared with analytical
predictions. Moment strengths of the connections were calculated using yield-line theory
to predict end-plate yielding and maximum bolt force calculations including prying
action. Experimental results were also compared to previous research with regards to
strength and stiffness. The inelastic rotation of connections was calculated and
conclusions were drawn on the compliance of these connections with current AISC
specifications.
Page 3
iii
Acknowledgements
I would like to express my deepest appreciation to my committee chairman, Dr.
Tom Murray. His guidance, patience and wisdom over the course of this research was
invaluable. I would like to thank Dr. Sam Easterling and Dr. Thomas Cousins for serving
as committee members. I would also like to thank Dr. Richard Barker for serving as an
interim member of this committee at my defense while Dr. Easterling was on sabbatical.
I was very fortunate to have help from these people: Mike Bryant, Brian Diaz,
Marc Graper, Michelle Rambo-Roddenberry, Joe Howard, Ron Meng, Tim Mays and last
but not least, John Ryan, who helped me in ways too numerous to count. I would also
like to extend my deepest appreciation to Brett Farmer and Dennis Huffman for doing so
many things so well for “the Yankee”. Special thanks to Clark Brown for the many bolts
he instrumented for me, and to the Structures and Materials Program secretary, Ann
Crate.
I would like to thank the many friends I have here in Blacksburg and at home in
Wisconsin for helping me keep things in perspective and adding many moments of levity
during the course of this research. Finally, I would like to thank the two people without
whom none of this would have been imaginable, much less possible: my mom and dad.
Their support throughout the course of my graduate work was more than anyone could
possibly hope for.
Page 4
iv
TABLE OF CONTENTS
PageABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii
LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi
LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
CHAPTER
I. INTRODUCTION AND LITERATURE REVIEW . . . . . . . . . . . . . . . . . 1
1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Previous Research on Flush End-Plate Moment Connections. . . . . . . . . 21.3 Previous Cyclic Testing of End-Plate Moment Connections . . . . . . . . 31.4 Inelastic Rotation Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51.5 Objective and Scope of Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
II. Flush End-Plate Moment Connection Design . . . . . . . . . . . . . . . . . . . . . . . 8
2.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82.2 End-Plate Yield-Line Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82.3 Prediction of Bolt Forces Including Prying Action. . . . . . . . . . . . . . . . 15
III. TESTING PROGRAM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.1 Testing Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233.2 Test Set-up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233.3 Instrumentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283.4 Loading Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
IV. EXPERIMENTAL RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.1 Two-Bolt Flush End-Plate Tests . . . . . . . . . . . . .. . . . . . . . . . . . . 354.1.1 Comparison of Predicted and Tested Moment Strengths . . . . . 354.1.2 Rotation Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374.1.3 Condition at End of Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
4.2 Four-Bolt Flush End-Plate Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444.2.1 Comparison of Predicted and Tested Moment Strengths . . . . . 444.2.2 Rotation Capability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464.2.3 Condition at End of Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.3 Comparison With Previous Monotonic Load Tests . . . . . . . . . . . . . . . 50
Page 5
v
V. SUMMARY AND CONCLUSIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
APPENDIX A – NOMENCLATURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
APPENDIX B – TWO-BOLT FLUSH RESULTS AND TEST DATA . . . . . . . . . . 71
APPENDIX C – FOUR-BOLT FLUSH RESULTS AND TEST DATA . . . . . . . . . 99
Page 6
vi
LIST OF FIGURES
Figure Page
1.1 Examples of Flush End-Plate Configurations . . . . . . . . . . . . . . . . . . . . . . . 2
2.1 Yield Line Pattern for Two-Bolt Flush End-Plate . . . . . . . . . . . . . . . . . . . 13
2.2 Yield Line Pattern for Four-Bolt Flush End-Plate . . . . . . . . . . . . . . . . . . . 14
2.3 Kennedy Split-Tee Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.4 Modified Kennedy Model for Two-Bolt Flush End-Plate . . . . . . . . . . . . . . 18
2.5 Modified Kennedy Model for Four-Bolt Flush End-Plate . . . . . . . . . . . . . 18
3.1 End-Plate Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.2 Section Throught Test Set-Up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.3 Photograph of Test Set-Up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
3.4 ATC-24 Loading Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.5 SAC Loading Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
4.1 Schematic of Test Set-Up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.2 Basis of Reported Rotation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
4.3 Typical Moment Versus Total Rotation Curve . . . . . . . . . . . . . . . . . . . . . . . 40
4.4 Typical Moment Versus Inelastic Rotation Curve . . . . . . . . . . . . . . . . . . . . 40
4.5 Photograph of F1-3/4-3/8-16 Test 1 at Failure . . . . . . . . . . . . . . . . . . . . . . . 42
4.6 Photograph of Fracture at Flange-to-End-Plate Interface F1-3/4-3/8-16 . . . 43
4.7 Photograph of F1-3/4-3/8-16 Test 2 at Failure . . . . . . . . . . . . . . . . . . . . . . . 43
4.8 Web-to-End-Plate Weld Fracture F2-3/4-3/8-16 . . . . . . . . . . . . . . . . . . . . . . 49
4.9 Photograph of End-Plate Fracture F2-3/4-3/8-16 . . . . . . . . . . . . . . . . . . . . . 49
Page 7
vii
4.10 Photograph of Flange-to-End-Plate Weld Fracture F2-3/4-3/8-16 . . . . . . . . 50
4.11 Typical Moment Versus Total Rotation Curve With Superimposed Monotonic
Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
4.12 Comparison With Srouji (1983) F1-3/4-3/8-16 Test 1 . . . . . . . . . . . . . . . . . 54
4.13 Comparison With Srouji (1983) F1-3/4-3/8-16 Test 2 . . . . . . . . . . . . . . . . . 54
4.14 Comparison With Srouji (1983) F1-5/8-3/8-16 Test 1 . . . . . . . . . . . . . . . . . 55
4.15 Comparison With Srouji (1983) F1-5/8-3/8-16 Test 2 . . . . . . . . . . . . . . . . . 55
4.16 Comparison With Srouji (1983) F2-3/4-3/8-16 . . . . . . . . . . . . . . . . . . . . . . 56
4.17 Comparison With Srouji (1983) F2-5/8-3/8-16 Test 1 . . . . . . . . . . . . . . . . . 56
4.18 Comparison With Srouji (1983) F2-5/8-3/8-16 Test 2 . . . . . . . . . . . . . . . . . 57
Page 8
viii
LIST OF TABLES
Table Page
3.1 End-Plate Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.2 SAC Loading Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
4.1 Two-Bolt Flush Test Results – Strength Data . . . . . . . . . . . . . . . . . . . . . . 36
4.2 Two-Bolt Flush Test Results – Rotation Data . . . . . . . . . . . . . . . . . . . . . . 38
4.3 Four-Bolt Flush Test Results – Strength Data . . . . . . . . . . . . . . . . . . . . . . 45
4.4 Four-Bolt Flush Test Results – Rotation Data. . . . . . . . . . . . . . . . . . . . . . . 46
4.5 Comparison of Results With Srouji and Kline Prior to Adjustment . . . . . . 53
4.6 Comparison of Results With Srouji and Kline After Adjustment . . . . . . . . 58
5.1 Compliance With Inelastic Rotation Requirements . . . . . . . . . . . . . . . . . . 62
Page 9
1
CHAPTER I
INTRODUCTION AND LITERATURE REVIEW
1.1 INTRODUCTION
The moment end-plate connection is one of the fully restrained moment
connections as defined by the American Institute of Steel Construction (AISC) Manual of
Steel Construction, Load and Resistance Factor Design (LRFD) (Manual 1994). Bolted,
flush end-plate connections have been used extensively in the metal building industry as
splices between beams, and beam-column connections in portal frame construction. The
popularity of these connections stems from their ease of fabrication and erection in the
field.
The beam-to-column flush moment end-plate connection consists of a steel plate
welded to the end of a beam flange. The end-plate is then bolted to the column flange
using various rows of fully tensioned high-strength bolts. The flush end-plate moment
connection is one in which the end-plate does not extend above the beam flange. One or
two rows of bolts at each flange can be used. Figure 1 shows the two and four-bolt flush
configurations. This study is limited to these two configurations.
During the 1994 Northridge earthquake, many weld cracks were found in flange-
welded moment connections. Because of this, bolted moment connections have seen a
rise in popularity as engineers seek alternatives to the direct welded connection. The
flush moment end-plate is considered a fully restrained connection but as of yet has not
Page 10
2
been approved for usage in seismic zones due to lack of research into its inelastic rotation
capability.
The purpose of this research is to investigate the usage of the flush moment end-
plate connection in seismic areas. Experimental research was conducted to determine the
inelastic rotation capability of the flush moment end-plate connection, and to determine
its effectiveness as a fully restrained connection under seismic loading.
1.2 PREVIOUS RESEARCH ON FLUSH END-PLATE MOMENT CONNECTIONS
Zoetermeijer (1981) analyzed an infinitely long plate bounded by two fixed edges
and one free edge, and loaded with a concentrated force using yield-line theory. He
a) Two-Bolt Flush b) Four-Bolt Flush
FIGURE 1.1 Flush End-Plate Configurations
Page 11
3
developed a chart which may be used to calculate the ultimate load of a stiffened column
flange or a flush end-plate once the distance from the bolt to the beam flange and web are
known. He noted, however, that the analysis was only and approximation.
Phillips and Packer (1981) conducted a series of tests to determine the influence
of the end-plate thickness on moment-rotation characteristics and the end-plate collapse
mechanism. They also studied end-plate connections with two rows of tension bolts in
order to study the influence of the second row on the stiffness of the connection. They
suggested two failure mechanisms for end-plates with two rows of bolts, by which they
determine the required thickness of the end-plate. They concluded that flush end-plates
with two rows of bolts in the tension region are suitable for semi-rigid construction, and
that the second row of tension bolts is effective but to a much lesser extent than
previously estimated.
Srouji et al. (1983) investigated the behavior of four types of end-plate moment
connections including the two-bolt, and four-bolt flush connection included in this study.
Srouji investigated the use of yield-line analysis for predicting the failure moment, and
the Kennedy method, modified for use with end-plates, to predict bolt forces for these
connections. Srouji used a combination of experimental results, finite element analysis,
and comparison with previous research to verify the analytical predictions.
Hendrick et al. (1985) performed more experiments on flush end-plates and
developed a unified design methodology for flush end-plates. Hendrick developed an
empirical equation to estimate the distance from the bolt line to the location of prying
forces.
Page 12
4
Borgsmiller and Murray (1995) performed experiments on flush, extended
stiffened, extended unstiffened, and multi-row extended end-plate moment connections.
They devised a simplified method for design of end-plate moment connections based on
two limit states, end-plate yielding, and bolt rupture. Yield line analysis was used to
predict the strength due to end-plate yielding, and a modified version of the Kennedy
method was used to calculate the connection strength due to bolt-rupture. Proposed in
the study was a simplified method to calculate connection strength due to bolt rupture.
This method has is origin in the Kennedy method, but a simplifying assumption is made
to greatly reduce the time and complexity of calculation involved in using the Kennedy
method.
1.3 PREVIOUS CYCLIC TESTING OF END-PLATE MOMENT CONNECTIONS
The tests discussed in the previous section were all performed under static loading
conditions. Up to this point, there has been very little research done on the performance
of flush end-plate moment connections under seismic loading. Research on the
performance of four-bolt-extended-unstiffened and four-bolt-extended-stiffened end-plate
moment connections under cyclic loading was performed by Tsai and Popov (1990).
While they only did a limited number of tests, Tsai and Popov found the four-bolt-
extended-unstiffened end-plate connection to perform well under cyclic loading. They
did however have to increase bolt and end-plate sizes over what was required by the
current design methods at the time.
Page 13
5
Ghobarah et al. (1990) investigated extended end-plates under cyclic loading to
investigate the effects of design parameters such as end-plate thickness, column flange
stiffener and bolt pre-tension force on the overall behavior. They found that end-plate
connections are able to dissipate energy induced in the connection from cyclic loading
without loss of strength. This energy dissipation can be achieved through proper
detailing of the connection.
Meng (1996) addressed the issue of seismic design of end-plate moment
connections. Through the use of full scale moment end-plate connection tests, tee stub
tests, and finite element modeling, Meng developed an LRFD based design procedure for
four-bolt extended, shimmed four-bolt, and 8-bolt (four-bolt wide) moment end-plate
connections. The presence of weld access holes in some of the tests were determined to
be the cause of flange fractures during these tests. He deemed the four-bolt extended
end-plate moment connection satisfactory for use in seismic zones. He also indicated that
the 8-bolt (four-bolt wide) connection exhibits satisfactory performance under cyclic
loading, but also reports that more research and analysis is needed before this connection
can be considered as an alternative to flange welded connections.
Kline et al. (1989) investigated the performance of several end-plate
configurations, including flush end-plate connections, under wind loads, subjecting them
to numerous cycles. Kline et al. concluded that end-plate moment connections using
snug tight bolts are adequate in resisting expected wind loads, but provide reduced
stiffness when compared to moment end-plates with fully tightened bolts.
Page 14
6
1.4 INELASTIC ROTATION REQUIREMENTS
Current design recommendations (FEMA 1995) and the AISC Seismic Provisions
for Structural Steel Buildings (Seismic 1997) divide moment frames into three broad
categories: special, intermediate, and ordinary. Beam-to-column connections used in
each of the three categories must meet strength and inelastic rotation requirements. Only
tested connections are permitted for use in the Seismic Force Resisting System.
Special moment frames are expected to withstand significant inelastic
deformations during the design seismic event and their beam-to-column connections must
exhibit an inelastic rotation of at least 0.03 radians. Beam-to-column connections shall
demonstrate a flexural strength at the column face of at least the plastic moment strength
of the beam. When beam local buckling rather than beam yielding limits the flexural
strength of the beam or when reduced flange width beam sections are used, the
connections are allowed to be designed to withstand 0.8Mp of the tested beam.
Connections that provide the required rotation through deformation of the various
connection elements while maintaining the required design strength are permitted as long
as the effects on drift and overall frame stability are considered.
Intermediate moment frames are expected to withstand moderate inelastic
deformations during the design seismic event and their beam-to-column connections must
exhibit an inelastic rotation of at least 0.02 radians. Fully-restrained (FR) connections in
an intermediate moment frame must attain the required inelastic rotation through the
plastic deformation of members in the frame. Partially-restrained (PR) connections are
allowed for usage in an intermediate moment frame and may obtain the required inelastic
Page 15
7
rotation through the plastic deformation of the connections themselves, such as plate
yielding and bending of bolts. As with special moment frames, beam-to-column
connections must have at least the same moment strength at the column face as the plastic
moment strength of the beam. Like special moment frames, a connection strength of
0.8Mp is permitted if beam local buckling is the controlling limit state, or if a reduced
beam section is used.
Ordinary moment frames are expected to withstand limited inelastic deformations
during the design seismic event and their beam-to-column connections must exhibit an
inelastic rotation of at least 0.01 radians. FR connections in the seismic force resisting
system must be designed for a moment strength of at least 1.1RyMp of the beam or the
maximum moment that can be delivered to the system, whichever is larger. The
Parameter Ry is the ratio of the expected yield strength to the minimum specified yield
strength. PR connections are also permitted for use as long as certain strength and frame
stability conditions are met.
1.5 OBJECTIVE AND SCOPE OF RESEARCH
The primary purpose of this research is to investigate the performance of flush
end-plate moment connection under seismic loading. Seven full-scale tests were
conducted to determine the inelastic rotation capability of these connections. The
specimens were designed using a combination of yield-line theory and maximum bolt
force predictions which include an estimate of prying forces. The design methodology
for flush end-plate moment connections is described in Chapter II. The test specimens,
Page 16
8
set-up, and procedures are described in Chapter III. Results of the full-scale testing is
given in Chapter IV, with summary, conclusions and recommendations presented in
chapter V.
Page 17
9
CHAPTER II
FLUSH END-PLATE MOMENT CONNECTION DESIGN
2.1 OVERVIEW
The flush end-plate moment connections in this study were designed taking into
account the two limit states of end-plate yielding and bolt rupture. The strength of the
connection due to end-plate yielding was found using yield line theory, and the
connection strength due to bolt rupture was found using a modified version of the
Kennedy method including prying forces.
Included in this chapter is the development of the yield line mechanisms to
calculate the connection strength due to end-plate yielding, and the development of the
modified Kennedy method to calculate the connection strength due to bolt rupture.
2.2 END-PLATE YIELD-LINE STRENGTH
Yield-line theory was first introduced as a way to analyze the behavior of
concrete slabs. A yield line is a continuous formation of plastic hinges along a straight or
curved line. A failure mechanism is said to exist when the yield lines form a
kinematically valid collapse mechanism. The following guidelines are followed when
establishing yield line patterns:
1. Axes of rotation generally lie along lines of support.
2. Yield lines pass through the intersection of the axes of rotation of adjacent
plane segments.
Page 18
10
3. Along every yield line, the bending moment is assumed to be a constant and is
taken as the plastic moment of the plate.
The following is a discussion of the yield-line analysis of flush end-plate
performed by Srouji et al (1983). The analysis of a yield-line mechanism can be
performed by two different methods, the equilibrium method and the virtual work or
energy method. Srouji et al. analyzed yield line mechanisms for flush end-plates using
the virtual work method. In this method, the external work done by the applied load from
a small arbitrary virtual deflection is set equal to the internal work done as the plate
rotates at the yield lines to accommodate this virtual deflection. For a specified yield-line
pattern and loading, a certain plastic moment will be required along the hinge lines. For
the same loading, other patterns may result in a larger required plastic moment capacity.
Hence, the controlling pattern is the one which requires the largest required plastic
moment. Or, for a given plastic moment capacity, the controlling mechanism is the one
which produces the lowest failure load. This implies that the yield-line theory is an upper
bound procedure and the least upper bound must be found.
To determine the required plastic moment capacity or the failure load, a number
of possible yield-line mechanisms must be investigated. By equating the internal and
external work, the relationship between the applied loads and the ultimate resisting
moment is obtained. The resulting equation is then solved for either the unknown loads
or unknown moments, and by comparing the different values obtained from the various
mechanisms, the controlling minimum load (or maximum required plastic moment) is
obtained.
Page 19
11
The internal energy stored in a particular yield-line mechanism is the sum of the
internal energy stored in each yield line forming the mechanism. The internal energy
stored in any given yield line is obtained by multiplying the normal moment on the yield-
line with the normal rotation of the yield-line. Thus the energy stored in the nth yield-
line of length Ln is:
where θn is the relative rotation of line n, and ds is the elemental length of line n. The
internal energy stored in a yield-line mechanism can be written as:
where N is the number of yield lines in the mechanism.
Srouji et al. investigated a number of yield line mechanisms for two-bolt and four-bolt
flush end-plates. They found that for each end-plate configuration one yield-line
mechanism controlled. The controlling yield line mechanism for the two bolt flush end-
plate is shown in Figure 2.1. The yield line mechanism for the four-bolt flush endplate is
shown in Figure 2.2. The geometric parameter are also defined in the figures. The
internal energy stored in the yield-line mechanism shown in Figure 2.1 is:
∫=nL npin dsmW θ
∑ ∫=
=N
1nL npi
n
dsmW θ
∑=
=N
1nnnp Lm θ
(2.1)
(2.2)
Page 20
12
FIGURE 2.1 Yield Line Pattern for Two-Bolt Flush End-Plate
a) Dimensions b) Yield Line Mechanism
Page 21
13
a) Dimensions b) Yield Line Mechanism
FIGURE 2.2 Yield Line Pattern for Four-Bolt Flush End-Plate
s
Page 22
14
where pf is the distance from the bolt centerline to the face of the flange, equal to pt – tf,
and s is the distance between parallel yield lines. The quantity s in Equation 2.3 is
obtained by differentiating the internal work equation with respect to s and equating to
zero, resulting in:
The internal energy stored in the straight yield-line mechanism for a four-bolt flush end-
plate shown in Figure 2.3 is given by:
where pb is the pitch between bolt rows, and s is the distance between parallel yield lines.
The quantity s is determined in the same manner as s in Equation 2.3 e.g., by
differentiating the internal work equation with respect to s and equating to zero, resulting
in:
( ) ( )
++
+=
g2
sps1
p1
2b
hp-h
4mW ff
ftpi
gb21s f=
( )( )t
btf p-h
p-p-hgb
21
s =
( ) ( )
−
+++
+=
2spb
sppg2
s1
p1
2b
p-hh
4mW bf
bff
ft
pi
(2.3)
(2.4)
(2.5)
(2.6)
Page 23
15
For all the end-plates, the external work done due to a unit displacement at the top of the
tension beam flange, resulting in a rotation of the beam flange, resulting in a rotation of
the beam cross-section about the outside of the compression flange is given by:
We = Muθ
Where Mu is the factored beam moment at the end-plate, and θ is the rotation at the
connection, equal to 1/h, where h is the beam depth.
On equating the internal and external work terms and canceling θ, the expression
for the ultimate moment, Mu, can be obtained. Then by rearranging the expression for
Mu, the equation for tp, the end-plate thickness, can be written in terms of Mu. To obtain
equations for Mu and tp in terms of Fpy and tp, mp can be replaced with the following:
where mp is the plastic moment strength of the end-plate per inch of width, Fpy is the
yield strength of the end-plate, and tp is the end-plate thickness. The resulting equations
for required end-plate thickness from those yield-line patterns are as follows.
For a two-bolt flush end-plate:
( )g2
sps1
p1
2b
FM
t
ff
f
py
u
p
++
+
= (2.9)
4tF
m2
ppyp =
(2.7)
(2.8)
Page 24
16
For a four-bolt flush end-plate:
From these equations connection strength according to yield-line theory can be
determined as follows:
For a two-bolt flush end-plate.
For a four-bolt flush end-plate:
2.3 PREDICTION OF BOLT FORCES INCLUDING PRYING ACTION
Kennedy, et al. (1981), used a split-tee analogy to predict bolt forces for tee-
hanger connections. In this study, three types of plate behavior was identified. The first
case, to which all plates belong under low applied load, is identified by the absence of
( ) ( )2spb
sppg2
s1
p1
2b
p-h
FM
tbf
bff
ft
py
u
p
−
+++
+
= (2.10)
( ) ( )
++
+=
g2
spp1
s1
2b
p-hFtM ff
ftpy
2ppl
( ) ( )
−
+++
+=
2spb
sppg2
s1
p1
2b
p-hFtM bfbf
f
ftpy
2ppl
(2.11)
(2.12)
Page 25
17
plastic hinges in the plate. Bolt prying force is assumed to be zero, and the plate is said
to be “thick”. The next stage that the plate passes through is an “intermediate” stage.
This is exemplified by a plastic hinge that forms at the flange. The plate is said to be of
intermediate thickness and the prying force is somewhere between zero and a maximum
possible value. As the load increases, a second plastic hinge forms at the bolt lines and
the prying forces are considered to be at a maximum. When the plate is in this stage it is
considered to be “thin”. The model for this theory is shown in Figure 2.3.
This model was applied to flush end-plate moment connections by Srouji et al..
For the flush two-bolt connection, the model is simply one half of the Kennedy Split-Tee
Figure 2.3 Kennedy Split-Tee Model
Page 26
18
Model and is shown in Figure 2.4. For a flush four-bolt connection, Kennedy’s Split-Tee
Model was modified. This model is shown in Figure 2.5.
To determine which stage of behavior the end-plate is in, the thick plate limit and,
if necessary, the thin plate limit must be calculated. Using a flange force calculated from
a pre-determined moment, the thick plate limit, t1, is calculated using the following
equation:
Where Ff = total flange force. Because this equation is iterative, the following
approximation is given to find a preliminary value for t1 with which to start iterations:
Once the thick plate limit is determined, it is compared to the actual end-plate thickness.
If tp > tl, then the end-plate is considered to be “thick” and the prying force is taken to be
zero. If tp < t1, then the prying force is not zero and the thin plate limit, t11, must be
calculated using the following equation:
2
1f
f2pyf
ff1
tbF
3-Fb
F4pt
=
pyf
ff1 Fb
F4.21pt =
2
11
f2py
2
11f
f2py
f
yb3
bff
11
tw2F
3Fwtb
F3-F
2b
16Fd
-pF2
t
′−′+
=
π
(2.13)
(2.14)
(2.15)
Page 27
19
FIGURE 2.5 Modified Kennedy Model for Four-Bolt Flush End-Plate
FIGURE 2.4 Modified Kennedy Model for Two-Bolt Flush End-Plate
Page 28
20
where w’ = width of end-plate per bolt line minus the diameter of the bolt. Because the
equation for t11 is iterative, an approximation is given to find a preliminary value for t11
with which to start iterations:
To insure that the end-plate is not failing due to shear effects, the following equation
should be satisfied prior to starting iterations with Equation 2.5.
Where t11 is found from Equation 2.16. If Equation 2.17 is not satisfied, the beam
capacity must be increased so that shear failure does not occur. The thin plate limit is
then compared to the actual end-plate thickness. If tp > t11, then the plate is considered to
be intermediate, the prying forces are not equal to zero, and can be calculated using the
appropriate equations. If tp < t11, then the end-plate is considered to be thin, the prying
forces are at a maximum and are calculated using the appropriate equations.
Kennedy et al. (1981) considered bolt forces to be the sum of a portion of the
flange force and the prying forces. The prying action varies according to what stage of
behavior the end-plate is in. For thin end-plates the prying force is at a maximum and is
given by the following equation:
′+
=w8.0
20.85b
F
16Fd
-pF2
tf
py
yb3
bff
11
π
3
Ftw2F py11
f
′<
(2.16)
(2.17)
Page 29
21
where Qmax = maximum prying force, a = distance from the prying force to the bolt line,
and F’ = flange force per bolt. Where F’ is taken as the lesser of Flimit or Fmax as
calculated by the following:
An equation for the calculation of the “a” distance was developed by Hendrick et
al. (1985) and verified by experimental results. This equation is a function of the end-
plate thickness and the bolt diameter and is given as:
Srouji et al. (1983) give two equations for prying forces of flush end-plates in the
intermediate stage, one for two-bolt flush, and one for four-bolt flush. The prying force
for an intermediate two-bolt end-plate is:
2
p
2py
2p
max twF
3f4atw
Q
′′−
′=
f
yb3
bfpy
2p
limit 4p8
Fdw0.80
20.85b
FtF
π+
′+
=
( )2
FFor maxf
max =
085.0dt
3.682a3
b
p −
=
(2.19)
(2.20)
(2.21)
(2.18)
Page 30
22
where F = flange force per bolt = Ff/2 and “a” is found from Equation 2.21. The prying
force equation given by Srouji et al. for an intermediate four-bolt flush end-plate is:
where F = Ff/2. For two-bolt flush end-plate connections, half of the flange force plus
prying forces is assumed to be distributed to each of the bolts .
For four-bolt flush end-plate connections, the amount of flange force distributed
to each bolt is dependent on the stage of end-plate behavior. For a thick end-plate, the
outer row of bolts is assumed to carry all of the flange force, while the force in the inner
row of bolts is assumed to be zero. For an intermediate end-plate, the outer bolt force, B,
is given by:
For a thin end-plate the outer bolt force, B, is given by:
32aFd
tb2F
3F8atb
aFp
Q yb2
b
2
2pf
2py
2pff
π−
−−=
Q5.2
FB f +=
maxf Q
83F
B +=
(2.22)
(2.23)
(2.24)
(2.25)
( )( ) ( )b
yb3
b
2
2pf
2py
b
pf
b
bf
pa16Fd
tb
2F3F
pa8tb
pap1.0pF
Q2
+−
−
+−
++
=π
Page 31
23
By setting the bolt force, (B), to the unfactored tensile strength of the bolt, these
equations can thus be solved for a flange force, which in turn can be used to calculate the
moment strength of the connection due to bolt fracture.
Page 32
24
CHAPTER III
TESTING PROGRAM
3.1 TESTING PROGRAM
Full scale testing was conducted using built up sections and flush end-plate moment
connections. A325 bolts and A572 Gr50 steel were used for the connection components.
Two flush end-plate configurations were tested in the study: 2-bolt and 4-bolt
unstiffened. The nominal geometry is listed in Table 3.1,and the details of the end-plate
connections tested are shown in Figure 3.1. The test designations shown in the Table 3.1
are to be interpreted as follows: F1-3/4-3/8-16 designates a flush end-plate with one row
of ¾ in. diameter bolts at each flange, the end-plate thickness is 3/8 in. and the beam
depth is 16 in. Because the testing was cyclic, the same bolt pattern was used at both
flanges. The connections were designed so that failure would occur within the
connection elements and not in the test beam or column. This was achieved by
computing an end-plate strength from yield line theory and the bolt force prediction
method described in Chapter II, and then designing beams and columns to be
approximately ten percent stronger than the connections. The same end-plate was welded
to both sides of the test beam allowing two tests for each end-plate configuration.
Page 33
25
3.2 TEST SET-UP
The physical test setup for the evaluation of the connections was a cantilevered
beam connected to a column section as shown in Figures 3.2 and 3.3. A section cut
through the test set-up is shown in Figure 3.2. The test setup was erected upright in the
vertical plane, eliminating any moment caused by gravity forces
TestDesignation
BoltDiameter
(in)
End-PlateThickness
(in)
BeamDepth(in)
FlangeWidth(in)
Pitch(in)
Gage(in)
F1-3/4-3/8-16 3/4 3/8 16 6 1 1/2 3 1/2
F1-5/8-3/8-16 5/8 3/8 16 6 1 3/8 2 3/4
F2-3/4-3/8-16 3/4 3/8 16 6 1 1/2 3 1/2
F2-5/8-3/8-16 5/8 3/8 16 6 1 3/8 2 3/4
perpendicular to the loading plane. The test column was bolted to the test set-up in a way
to create what was essentially a simple connection. Whether or not this in fact was a
truly simple connection was not a concern in this study, because the rotation was
measured in a manner that accounts for both rigid body motion of the test frame and test
column rotation at the connection. Lateral bracing of both beam flanges was provided at
intervals such to prevent failure by lateral torsional buckling.
TABLE 3.1 End-Plate Parameters
Page 34
26
FIGURE 3.1 End-Plate Details
Page 35
27
FIGURE 3.2 Section Through Test Set-Up Diagram
Page 36
28
FIGURE 3.3 Test Setup
Page 37
29
The specimen was loaded by means of a 200 kip hydraulic ram placed 6 ft from the
center of the test column. No axial loads were applied to the column or beam sections.
3.3 INSTRUMENTATION
Instrumentation for recording of test data included a 200 kip capacity tension-
compression load cell, a linear displacement transducer to measure beam deflection at a
distance of 1 ft offset from the load point, potentiometers to measure the end-plate and
column flange separation at the top and bottom of the end-plate, strain gauges to measure
beam flange strain and instrumented bolts to measure bolt strain. Linear transducers were
also used to measure the vertical displacement and rotation of the test column. This
allowed for the determination of the rotation at the end-plate, taking into account the
rotation and vertical displacement of the test column.
The end-plate connection bolts used in the tests were instrumented with an
internal strain gauge to measure real time strain within each bolt. In addition to
recording bolt strain during these tests, the instrumented bolts allowed for accurate
determination of pre-tension force during tightening.
To prepare the instrumented bolts, a 2 mm hole was drilled in the head of the bolt
to a depth such that a special strain gauge could be inserted below the head of the bolt but
above the threaded portion of the shank. This is done such that the 2 mm hole does not
reduce the strength of the bolt. This is possible because the reduction of the bolt shank
area is much less than the reduction of area in the threaded portion of the bolt. After
insertion of the strain gauge, an epoxy was injected into the hole which, upon curing,
Page 38
30
formed a tight bond between the gauge and bolt material. Each bolt was then calibrated
for tension load versus strain using a tensile test machine. Data from each test was
recorded from test setup to eventual failure, through the use of the System 4000 data
acquisition system. The data was transferred via disk media to commercial software for
analysis in spreadsheet and graphical analysis software.
3.4 LOADING PROTOCOL
The specimens were tested using one of two loading protocols. The first three
tests (F1-3/4-3/8-16 tests 1 and 2, and F2-3/4-3/8-16) were conducted using the loading
protocol outlined in ATC-24, Guidelines for Cyclic Seismic Testing of Components of
Steel Structures (1992). The procedure is for use in slow cyclic load application which is
significantly less than the real time frequency of an earthquake. This method of testing is
used for its cost effectiveness and allows for more observation of damage to the structural
components as they occur. A major limitation of this method is that it severely distorts
time and its material property effects. ATC-24 states, however, that results from tests
using this methodology can be considered conservative because the slow cyclic loading
results in a small decrease in strength and an increase in the rate of deterioration.
Although this may seem to limit the value of the results, slow cyclic testing does provide
valuable data for overall connection performance.
To gain an understanding of the nature and process of cyclic testing of
connections, the following definitions are given from ATC-24.
Page 39
31
Cycle: a load or deformation history unit consisting of two sequential
excursions, one in the positive and one in the negative loading
direction.
Deformation: a generic quantity, δ, including strains, angles of shear distortion,
rotations, axial deformations and displacements.
Ductility Ratio: a ratio of peak deformation over yield deformation.
Excursion: a load or deformation history unit that starts and finishes at zero load,
and contains a loading and unloading branch.
Force: a generic quantity, Q, including internal forces and externally applied
loads.
Hysteretic Area: the area enclosed by a force-deformation diagram.
Load or Deformation Step: a load history unit consisting of a series of cycles with
constant peak load or deformation.
Peak Deformation: the deformation at a load reversal point.
Total Deformation Range: the total deformation between the peak of an excursion and
the peak of the preceding opposite excursion.
Yield Force of Deformation: the predicted or measured force or deformation at which
significant yielding occurs.
While using the ATC-24 loading protocol two methods of loading control are
utilized during each specimen test, force and deformation. As their names imply, the
control of loading by the tester is either by the force applied to the specimen or by the
measured deformation of the specimen. For the tests performed in this study, the force
Page 40
32
control is monitored from values given by the load cell attached to the hydraulic ram used
to apply load. The deformation method is in reference to the deflection of the specimen
at a distance inboard one foot from the load point.
The ATC-24 loading history is shown as Figure 3.4. ATC-24 makes several
recommendations on the numbers of cycles and peak deformations in each load step.
They are as follows:
• The number of cycles no with a peak deformation less than δy should be at least six.
• The number of cycles n1 with a peak deformation δ1 equal to δy should be at least
three.
• The number of cycles n2 with peak deformation δ2 = δy + ∆ should be at least three
unless a lower number can be justified.
• The number of cycles n3 with peak deformation δ3 = δy + 2∆ should be at least three
Load Steps-5
-4
-3
-2
-1
0
1
2
3
4
5
Number of Cycles
Deformation
1 2 3 4 5 6 7
δy
FIGURE 3.4 ATC-24 Loading Protocol
Page 41
33
unless a lower number can be justified.
• The number of cycles n4 to nm with peak deformation δ4 = δy +3∆ to δm = δy + (m –
1)∆ should be at least two unless a lower number can be justified.
To use this loading protocol, an accurate calculation of the yield load of the
specimen is needed. All of the cycles with a peak deformation below the yield
deformation were performed using load control. Once the specimen reaches the yield
load, deformation control is used exclusively. In this study, the specimen was loaded to a
force, Q, equal to 0.5Qy for the first three cycles, then the specimen was loaded to a force,
Q, equal to 0.75Qy for the second three cycles, where Qy is equal to the yield load of the
connection. Once the specimen reached the yield load, the deformation was recorded.
This then became the yield deformation, δy, described in ATC-24. This deformation also
becomes the ∆ that the specimen is loaded in multiples of for the rest of the experiment.
The analytical calculation of the yield deflection needed for use of the ATC-24
loading protocol proved to be somewhat problematic. Therefore, the remaining four tests
(F1-5/8-3/8-16 tests 1 and 2, and F2-5/8-3/8-16 tests 1 and 2) were performed using the
loading protocol defined by the SAC Protocol for Fabrication, Inspection, Testing, and
Documentation of Beam-Column Connection Tests and Other Experimental Specimens
(1997). This will be hereafter defined as the SAC Protocol. (SAC is a joint venture
involving the Structural Engineers Association of California (SEAOC), Applied
Technology Council (ATC), and the California Universities for Research in Earthquake
Engineering (CUREe)). Like the ATC-24 loading protocol, the SAC protocol is a
multiple step test. The main difference between the ATC-24 and the SAC protocol is that
Page 42
34
the SAC protocol is completely deformation controlled. The deformation parameter used
to control the loading history is the interstory drift angle, θ, defined as interstory
displacement divided by the story height. In the test specimen, this angle is defined as
the beam deflection divided by the beam span (to the column centerline) if the vertical
beam deflection is controlled, or as the column deflection divided by the column height if
the horizontal column deflection is controlled. As in the ATC-24 loading protocol, the
cycles are symmetric in peak deformations. The history is divided into steps and the peak
deformation of each step, j, is given as θj, a predetermined value of the interstory drift
angle. The SAC protocol proved far easier to use than the ATC-24 protocol because the
values of θj remain the same for all experiments, making it unnecessary to calculate the
“first yield” deflection of the specimen. The SAC protocol is shown graphically in
Figure 3.5, and in tabular form in Table 3.2.
-0.06
-0.05
-0.04
-0.03
-0.02
-0.01
0
0.01
0.02
0.03
0.04
0.05
0.06
Number of cycles
Inte
rsto
ry D
rift
Ang
le
6 6 6 4 2 2 2 2 2
FIGURE 3.5 SAC Loading Protocol
Page 43
35
Load Step # Peak Deformation, θ Number of Cycles, n
1 0.00375 6
2 0.005 6
3 0.0075 6
4 0.01 4
5 0.015 2
6 0.02 2
7 0.03 2
Continue with increments in θ of 0.01 rad., and perform two cycles at each step
Results, including comparison of predicted and tested moment strengths, total and
inelastic rotation capabilities, and condition at the end of the test are presented in Chapter
IV. Also included is a comparison of the current study with previous research.
TABLE 3.2 SAC Loading Protocol
Page 44
36
CHAPTER IV
EXPERIMENTAL RESULTS
4.1 TWO-BOLT FLUSH TESTS
4.1.1 COMPARISON OF PREDICTED AND ACTUAL MOMENT STRENGTHS
Four tests were conducted on two different flush two-bolt end-plate
configurations. Two tests were performed on the two-bolt configuration with 3/4 in.
bolts (F1-3/4-3/8-16), and two tests were performed on the two-bolt configuration with
5/8 in. bolts (F1-5/8-3/8-16). Gage and pitch for these connections are given in Table
3.1. A nominal design yield stress of 55 ksi was used for design of the connections. The
connections were designed without weld access holes.
Quasi-static or “slow cyclic” loading was applied to the cantilevered beam using a
hydraulic actuator. The loading history prescribed by ATC-24 and presented in Chapter
III was used for the two F1-3/4-3/8-16 tests. The loading history prescribed by SAC and
presented in Chapter III was used for the two F1-5/8-3/8-16 tests. Detailed presentations
of the data for all of the tests are located in Appendix B.
The F1-3/4-3/8-16 tests used instrumented bolts, allowing them to be tightened to
the minimum bolt tension of 28 kips as prescribed by the AISC LRFD Design
Specification for Structural Steel Buildings (Load 1993). Instrumented bolts were not
used for the F1-5/8-3/8-16 tests because the shanks of the instrumented bolts originally
intended for use in the test were too long, requiring the use of an excessive amount of
Page 45
37
washers. Uninstrumented bolts were used for these tests and tightened according to the
“turn of the nut” method.
The maximum moment strength of the connections from the tests correlated
poorly with the analytical strength predictions from yield-line analysis and bolt force
predictions including prying action. The methods used for the analytical predictions are
described in Chapter II. Comparison of predicted versus tested failure moments are given
in Table 4.1. In Table 4.1, Mpl is the moment strength of the connection calculated using
yield line theory, and Mq is the moment strength of the connection calculated using bolt
force predictions including prying action. The shaded boxes are the predicted controlling
limit state for the connection. The end-plate yield stresses, (Fpy), from coupon tensile
tests are shown in the last column. All yield stresses are approximately 53 ksi for the
end-plates.
Maximum Moment (ft-kips)
Predicted StrengthsConnection Test
Mpl Mq
AppliedMpl/Mapp Mq/Mapp
Fpy
(ksi)
F1-3/4-3/8-16 11 51.5 78.0 44.7 1.15 1.74 53.6
F1-3/4-3/8-16 21 50.0 78.3 44.5 1.12 1.76 52.6
F1-5/8-3/8-16 12 56.2 57.1 43.6 1.29 1.31 53.8
F1-5/8-3/8-16 22 60.4 58.0 45.3 1.33 1.28 53.5
1 ATC Loading Protocol used to load specimen.2 SAC Loading Protocol used to load specimen.
Table 4.1 Two-Bolt Flush Test Results – Strength Data
Page 46
38
The condition at the end of the tests are reported instead of the failure mode.
Because the rotation applied to the connections in this study is far beyond what is
normally required for other than earthquake loads, the failure mode or limit state does not
have the same significance as for monotonic tests. For an end-plate connection, if
enough rotation is applied, the connection will exhibit either bolt fractures, weld fractures
or end-plate fractures. Because of this, the yield-line solution can not be used to predict
the final condition of the specimen, it can only predict the moment strength of the
connection.
The yield-line predicted strengths were calculated using actual measurements of
gage and pitch of the connections and end-plate yield stress from tensile coupon tests.
Nominal bolt material strength was used to determine Mq. The tested maximum moment
was taken as the maximum moment at the end-plate that was resisted by the connection
during the entire test. For F1-3/4-3/8-16 tests 1 and 2 the moment strength given by yield
line theory is the controlling limit state but overpredicts by 15% and 12%, respectively.
For F1-5/8-3/8-16 test 1 the moment strength given by yield line theory is the controlling
limit state but is unconservative by 29%. For F1-5/8-3/8-16 test 2 the moment strength
from bolt force predictions including prying action is the controlling limit state but is
unconservative 28%.
4.1.2 ROTATION CAPABILITY
Rotation results of the two-bolt tests are summarized in Table 4.2. All rotations given
Page 47
39
in Table 4.2 are rotations at the end-plate. The number of completed cycles refers to the
number of cycles of the loading protocol that were completed prior to termination of the
test. Rotations were calculated by taking the deflection of the test beam at a point
inboard 1 ft from the applied load, dividing that deflection by the distance to the face of
the column and subtracting the measured rotation of the test column and the measured
rigid body motion of the load frame. A schematic of the test set-up is shown in Figure
4.1. The net or reported rotation is shown schematically in Figure 4.2.
Inelastic rotation was calculated using a simple linear interpolation from the
moment versus total rotation curve from the test. Two points were chosen on the
unloading curve of one of the first few cycles. It is during these first cycles that the
Maximum RotationConnection Test Completed
CyclesTotal Inelastic
Condition at Endof Test
F1-3/4-3/8-16 1 281 0.041 0.038 End-PlateFractures
F1-3/4-3/8-16 2 171 0.045 0.041 End-PlateFractures
F1-5/8-3/8-16 1 262 0.016 0.0093 Bolt Fracture
F1-5/8-3/8-16 2 262 0.018 0.013 Bolt Fracture
1 ATC Loading Protocol2 SAC Loading Protocol
TABLE 4.2 Two-Bolt Flush Test Results – Rotation Data
Page 48
40
FIGURE 4.1 Schematic of Test Setup
FIGURE 4.2 Basis of Reported Rotation
Page 49
41
connection exhibits completely elastic behavior. An equation was then written for the
line joining these two points. This was used as the equation representing the elastic
deformation of the connection. This allowed the elastic deformation of the connection to
be calculated for any moment. The inelastic rotation, θi, was then found by simply
subtracting the calculated elastic rotation, θe, from the total rotation, θt, data from the
test using the equation given below:
θi = θt - θe
where θi is the inelastic rotation, θt is the total rotation, and θe is the calculated elastic
rotation. Typical total and inelastic rotation curves are shown as Figures 4.3 and 4.4
respectively.
F1-3/4-3/8-16 was designed with a “wide” configuration and F1-5/8-3/8-16 was
designed with a “tight” configuration. “Tight” configurations have smaller gages and
pitches than “wide” configurations. This helps explain why both F1-3/4-3/8-16 tests
exhibited much greater inelastic rotation capability than F1-5/8-3/8-16 tests 1 and 2. The
wide configuration allows for a more flexible connection with more end-plate yielding.
Bolt fracture occurred in F1-5/8-3/8-16 tests 1 and 2 without any significant end-plate
yielding, while the F1-3/4-3/8-16 tests showed significant end-plate yielding. The tight
configuration combined with the use of smaller bolts in the F1-5/8-3/8-16 tests led to less
end-plate yielding, bolt fracture, and therefore, less inelastic rotation capability.
Page 50
42
Moment at End-Plate vs. Total Rotation at End-PlateF1-3/4-3/8-16 Test 1
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.05 -0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04 0.05Rotation (radians)
Moment (ft-kips)
Moment At End-Plate vs. Inelastic Rotation at End-PlateF1-3/4-3/8 Test 1
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.05 -0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04 0.05Rotation (radians)
Moment (ft-kips)
FIGURE 4.4 Typical Moment Versus Inelastic Rotation Curve
FIGURE 4.3 Typical Moment Versus Total Rotation Curve
Page 51
43
4.1.3 CONDITION AT END OF TEST
The condition at the end of each test for the two-bolt connections is given in
Table 4.2. The F1-3/4-3/8-16 connections had end-plate fractures, and bolt fractures
occurred in the F1-5/8-3/8-16 tests.
All of the end-plate fractures occurred at the weld locations. This means that the
end-plate cracked at the interface of the end-plate and the weld. There were no true
fractures of the welds, that is, where cracks formed in the welds.
Specimen F1-3/4-3/8-16 test 1 at failure is shown in Figure 4.5. The deformation
of the end-plate at the end of the test is shown in the photograph. The top and bottom
flanges have “peeled” away from the face of the column and the end-plate has deformed
significantly at the center. From the photograph it is difficult to see the end-plate
fractures at the interface of the web-to-end-plate weld and the end-plate. They became
more evident when the connection was disassembled and inspected. The fractures started
at the inside of the flanges and emanated toward the center of the web, 3-3/8 in. from the
top flange and 2-1/4 in. from the bottom flange. At the outer edge of the top and bottom
flange, a crack formed at the interface of the flange and the flange-to-end-plate weld and
the flange peeled away from the end-plate. This fracture is shown in Figure 4.6 as
highlighted by the red oval. It is evident that a crack has formed and the flange has
peeled away from the end-plate.
Specimen F1-3/4-3/8-16 test 2 at failure is shown in Figure 4.7. The photograph
shows the deformation of the end-plate at the end of the test. As in F1-3/4-3/8-16 test 1,
the top and bottom flanges have peeled away from the face of the column and the end-
Page 52
44
plate has deformed significantly at the center. It is possible to see the end-plate fractures
at the interface of the end-plate and the web-to-end-plate weld in the photograph. The
fractures started at the inside of the flanges and emanated toward the center of the web, 3-
3/8 in. from the top flange and 2-3/8 in. from the bottom flange. As in F1-3/4-3/8-16 test
1, cracks formed at the interfaces of the top and bottom flanges and flange-to-end-plate
weld. This occurred at the edges of both flanges and the flanges “peeled” away from the
flange-to-end-plate weld.
Bolt rupture occurred at the end of F1-5/8-3/8-16 tests 1 and 2. Both tests
showed very little deformation of the end-plate prior to bolt fracture. The only noticeable
yielding of the end-plate was around the bolts. This was noticed by flaking of the
whitewash around the bolts.
FIGURE 4.5 F1-3/4-3/8-16 Test 1 at End of Test
Page 53
45
FIGURE 4.6 Fracture at Flange-to-End-Plate Interface – F1-3/4-3/8-16 Test 1
FIGURE 4.7 F1-3/4-3/8-16 Test 2 at End of Test
Page 54
46
In test 1, one bolt ruptured, and in test 2 both bolts ruptured, both at the tension side of
the connection.
4.2 FOUR-BOLT FLUSH TESTS
4.2.1 COMPARISON OF PREDICTED AND ACTUAL MOMENT STRENGTHS
Three tests were conducted on two different flush four-bolt end-plate
configurations. One test was performed on the four-bolt configuration with 3/4 in. bolts
(F2-3/4-3/8-16), and two tests were performed on the four-bolt configuration with 5/8 in.
bolts (F2-5/8-3/8-16). Gage and pitch for these connections are given in Table 3.1. A
nominal design yield stress of 55 ksi was used for design of the connections. The
connections were designed without weld access holes.
Quasi-static or “slow cyclic” loading was applied to the cantilevered beam using a
hydraulic actuator. The loading history prescribed by ATC-24 and presented in Chapter
III was used for the F2-3/4-3/8-16 test. The loading history prescribed by SAC and
presented in Chapter III was used for the second two tests, namely F2-5/8-3/8-16 tests 1
and 2. Detailed presentations of the data for all of the tests are located in Appendix C.
The F2-3/4-3/8-16 test used instrumented bolts, allowing them to be tightened to
the minimum bolt tension of 28 kips prescribed by the AISC LRFD Design Specification
for Structural Steel Buildings (Load 1993). Instrumented bolts were not used for the F2-
5/8-3/8-16 tests. This was because the shanks of the instrumented bolts originally
intended for use in the test were too long, requiring the use of an excessive amount of
Page 55
47
washers. Uninstrumented bolts were used for these tests and tightened according to the
“turn of the nut” method.
Comparison of predicted versus tested maximum moments are given in Table 4.2.
As in Table 4.1, Mpl is the moment strength of the connection calculated using yield line
theory, and Mq is the moment strength of the connection calculated using bolt force
predictions including prying action. The end-plate yield stresses, (Fpy), from coupon
tensile tests are shown in the last column. All yield stresses were approximately 53 ksi
for the end-plates.
Maximum Moment (ft-kips)
Predicted StrengthsConnection Test
Mpl Mq
AppliedMpl/Mapp Mq/Mapp
Fpy
(ksi)
F2-3/4-3/8-16 11 65.1 104.6 66.8 0.97 1.57 52.7
F1-5/8-3/8-16 12 73.0 76.8 67.7 1.08 1.13 53.0
F1-5/8-3/8-16 22 73.0 76.8 67.9 1.10 1.13 53.0
1 ATC Loading Protocol used to load specimen.2 SAC Loading Protocol used to load specimen.
The yield-line predicted strengths were calculated using actual measurements of
gage and pitch of the connections and end-plate yield stress from tensile coupon tests.
Table 4.3 Four-Bolt Flush Test Results – Strength Data
Page 56
48
Nominal bolt material strength was used to determine Mq. The tested maximum moment
was taken as the maximum moment at the end-plate that was resisted by the connection
during the entire test. For all three tests, the predicted limit state was from the yield-line
analysis of the end-plate. The correlation between the predicted and tested moment
strengths for the four-bolt connections was slightly better than for the two-bolt
connections. The moment strength calculated using yield-line theory was conservative
by 3% for F2-3/4-3/8-16 and, for F2-5/8-3/8-16 tests 1 and 2, overpredicted by 8% and
10% respectively.
4.2.2 ROTATION CAPABILITY
Results for rotation of the four-bolt tests are given in Table 4.4. All rotations
given in Table 4.4 are rotations at the end-plate. The number of completed cycles refers
to the number of cycles of the loading protocol that were completed prior to termination
of the test.
Maximum RotationConnection Test Completed
CyclesTotal Inelastic
Condition at Endof Test
F2-3/4-3/8-16 1 121 0.052 0.048 End-Plate andWeld Fractures
F2-5/8-3/8-16 1 262 0.028 0.020 Bolt Fracture
F2-5/8-3/8-16 2 282 0.033 0.018 Bolt Fracture
1 ATC Loading Protocol2 SAC Loading Protocol
TABLE 4.4 Four Bolt Flush Test Results – Rotation Data
Page 57
49
Rotations were calculated in the same manner as the two-bolt tests, by taking the
deflection of the test beam at a point inboard 1 ft from the applied load, dividing that
deflection by the distance to the face of the column and subtracting the rotation of the test
column and the rigid body motion of the load frame. The inelastic rotation was
calculated using the same method that was used for the two-bolt tests.
F2-3/4-3/8-16 was designed with a “wide” configuration and F2-5/8-3/8-16 was
designed with a “tight” configuration. This helps explain why the F2-3/4-3/8-16
connection exhibited greater inelastic rotation capability than the F2-5/8-3/8-16
connection. The wide configuration allows for a more flexible connection with more
end-plate yielding. Bolt fractures occurred in both F2-5/8-3/8-16 tests 1 and 2 without
any significant end-plate yielding, while the F2-3/4-3/8-16 connection showed significant
end-plate yielding. The tight configuration combined with the use of smaller bolts in the
F2-5/8-3/8-16 tests led to less end-plate yielding, bolt fracture, and therefore, less
inelastic rotation capability.
4.2.3 CONDITION AT END OF TEST
The condition at the end of the test for each four-bolt connection is given in Table
4.4. End-plate and weld fractures occurred in test F2-3/4-3/8-16. Bolt fracture occurred
in both F2-5/8-3/8-16 tests 1 and 2.
In the F2-3/4-3/8-16 test, the end-plate fractured at the interface of the end-plate
and the extra web-to-end-plate weld at the bottom flange. A photograph of the
Page 58
50
connection at the end of the test is shown in Figure 4.8. The extra web-to-end-plate weld
fractured on the same side as the previously described end-plate fracture and is shown in
Figure 4.9 (Extra web-to-end-plate weld refers to the fabrication where one side of the
web is welded to the end-plate the full depth of the web and the other side is only welded
approximately 6 in. from the inside of the flanges toward the center of the web). Cracks
were also found in the flange-to-end-plate welds, and at the bottom flange, the flange had
completely torn away from the end-plate on the same side that the end-plate fracture
occurred as shown in Figure 4.10.
Bolt fracture occurred in both F2-5/8-3/8-16 tests with very little deformation of
the end-plate. The only noticeable yielding of the end-plate was around the bolts, as
evidenced by flaking of the whitewash around the bolts.
FIGURE 4.8 Web-to-End-Plate Weld Fracture – F2-3/4-3/8-16
Page 59
51
FIGURE 4.9 End-Plate Fracture – F2-3/4-3/8-16
FIGURE 4.10 Flange-to-End-Plate Weld Fracture – F2-3/4-3/8-16
Page 60
52
4.3 COMPARISON WITH PREVIOUS MONOTONIC LOAD TESTS
The results of this study are now compared to results from studies performed by
Srouji et al.(1983) and Kline et al. (1989). Predicted moment strengths and actual tested
moment strengths for the three studies are presented in Table 4.5. It is evident that the
results of this study do not correlate well with the results from the previous two studies.
Current predictions for maximum strength of the connections correlate well with previous
results, however, current maximum applied moments are much lower than those from
Srouji et al. and Kline et al.. The ratios of predicted moment strength and maximum
applied moment (Mpred/Mapp) are all near 1.0 for the two previous studies, while they are
larger in the current study.
One possible explanation for the unconservative results of the current study is an
out-of-calibration load cell used to measure the applied load. To test this hypothesis,
moment versus deflection curves were compared with those from Srouji et al. (1983).
Monotonic moment versus deflection curves were created from the cyclic tests in this
study by taking points at maximum moment for each load step in the loading histories.. A
typical moment versus total rotation curve for the current study is shown in Figure 4.11.
The darkened line shows the monotonic moment versus rotation curve for the test. The
rotation can then be multiplied by any distance to obtain a deflection at the corresponding
moment.
Page 61
53
Klin
e
1.03
1.03
1.18
1.18
0.97
0.91
0.91
1.03
0.10
5
Srou
ji
1.01
1.01
0.96
0.96
0.94
0.95
0.95
0.97
0.02
5
Cur
rent
1.15
1.12
1.29
1.33
0.97
1.08
1.08
1.15
0.11
7
Klin
e
60.0
60.0
57.0
57.0
70.0
91.0
91.0
Ave
rage
Stan
dard
D
evia
tion
Srou
ji
54.0
54.0
64.8
64.8
73.2
85.5
85.5
Cur
rent
44.7
44.5
43.6
45.3
66.8
67.7
67.9
Klin
e
62.0
62.0
67.4
67.4
68.2
83.0
83.0
Srou
ji
54.3
54.3
62.0
62.0
68.8
81.6
81.6
Cur
rent
51.5
50.0
56.2
60.4
65.1
73.0
73.0
Test 1 2 1 2 1 1 2
Con
nect
ion
F1-3
/4-3
/8-1
6
F1-3
/4-3
/8-1
6
F1-5
/8-3
/8-1
6
F1-5
/8-3
/8-1
6
F2-3
/4-3
/8-1
6
F2-5
/8-3
/8-1
6
F2-5
/8-3
/8-1
6
Max
imum
Mom
ent (
ft-ki
ps)
Pred
icte
d St
reng
ths
App
lied
Mpr
ed/M
app
TA
BL
E 4
.5 C
ompa
riso
n of
Cur
rent
Res
ults
With
Sro
uji (
1983
), an
d K
line
(198
9) P
rior
to A
djus
tmen
t
Page 62
54
The rotations from the monotonic moment versus rotation curves for this study
are multiplied by a distance to make the defections obtained analogous to those from
Srouji et al. This method was used to create the moment versus deflection curves used to
compare the initial stiffnesses of the tests in the current study with those from Srouji et al.
Before the connections yield, the initial stiffness of the connections in the current study
and Srouji et al. should be very close. The initial stiffness is the relationship between
moment and deflection while the connection is still behaving elastically. Theoretically,
there should be no difference between the initial stiffness of the connections in the
Moment vs. Total Rotation at End-PlateF2-5/8-3/8-16 Test 1
-80
-60
-40
-20
0
20
40
60
80
-0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04Rotation (radians)
Moment (ft-kips)
FIGURE 4.11 Typical Moment Versus Total Rotation Curve WithSuperimposed Monotonic Curve
Page 63
55
current study and the initial stiffness of the connections in the previous study (Srouji et
al. 1983)..
The initial stiffnesses of the connections from the current study did not correlate
well with those from Srouji et al. Because all comparisons are about the same, an
adjustment was made in attempt to match the initial stiffnesses of current results with
Srouji et al. A common factor of 1.2 was found and adjustments were made to the
current results. The monotonic moment versus deflection curves for the current study
were again compared with Srouji et al., this time multiplying the moments from the
current study by 1.2. The factor of 1.2 was found to produce the best correlation between
the current study and Srouji et al.. The monotonic moment versus deflection curves for
the current study and Srouji et al. are shown as Figures 4.12 and 4.13 for F1-3/4-3/8-16
tests 1 and 2 respectively, Figures 4.14 and 4.15 for F1-5/8-3/8-16 tests 1 and 2
respectively, Figure 4.16 for F2-3/4-3/8-16, and Figures 4.17 and 4.18 for F2-5/8-3/8-16
tests 1 and 2 respectively. The adjusted curves correlate well with those developed by
Srouji et al..
Table 4.6 contains the comparison of predicted yield-line moment strengths and
adjusted moment strengths for the three studies.
Page 64
56
FIGURE 4.12 Comparison of Srouji (1983) and AdjustedF1-3/4-3/8-16 Test 1 Results
FIGURE 4.13 Comparison of Srouji (1983) and AdjustedF1-3/4-3/8-16 Test 2 Results
Moment At End-Plate vs. Total Deflection at End-Plate
0
15
30
45
60
75
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00Deflection (in)
Moment (ft-kips)
CurrentSrouji
Moment At End-Plate vs. Total Deflection at End-Plate
0
15
30
45
60
75
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00Deflection (in)
Moment (ft-kips)
CurrentSrouji
Page 65
57
FIGURE 4.14 Comparison of Srouji (1983) and Adjusted F1-5/8-3/8-16 Test 1 Results
FIGURE 4.15 Comparison of Srouji (1983) and Adjusted F1-5/8-3/8-16 Test 2 Results
Moment At End-Plate vs. Total Deflection at End-Plate
0
15
30
45
60
75
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00Deflection (in)
Moment (ft-kips)
CurrentSrouji
Moment At End-Plate vs. Total Deflection at End-Plate
0
15
30
45
60
75
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Deflection (in)
Moment (ft-kips)
CurrentSrouji
Page 66
58
FIGURE 4.16 Comparison of Srouji (1983) and AdjustedF2-3/4-3/8-16 Results
FIGURE 4.17 Comparison of Srouji (1983) and Adjusted F2-5/8-3/8-16 Test 2 Results
Moment At End-Plate vs. Total Deflection at End-Plate
0
10
20
30
40
50
60
70
80
90
100
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00Deflection (in)
Moment (ft-kips)
CurrentSrouji
Moment At End-Plate vs. Total Deflection at End-Plate
0
10
20
30
40
50
60
70
80
90
100
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0Deflection (in)
Moment (ft-kips)
CurrentSrouji
Page 67
59
FIGURE 4.18 Comparison of Srouji (1983) and Adjusted F2-5/8-3/8-16 Test 2 Results
Moment At End-Plate vs. Total Deflection at End-Plate
0
10
20
30
40
50
60
70
80
90
100
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00Deflection (in)
Moment (ft-kips)
CurrentSrouji
Page 68
60
TA
BL
E 4
.6 C
ompa
riso
n of
Res
ults
With
Sro
uji (
1983
), an
d K
line
(198
9) A
fter A
djus
tmen
t
Klin
e
1.03
1.03
1.18
1.18
0.97
0.91
0.91
1.03
0.10
5
Srou
ji
1.01
1.01
0.96
0.96
0.94
0.95
0.95
0.97
0.02
5
Cur
rent
A
djus
ted
0.96
0.94
1.07
1.11
0.81
0.90
0.90
0.96
0.09
7
Klin
e
60.0
60.0
57.0
57.0
70.0
91.0
91.0
Ave
rage
Stan
dard
D
evia
tion
Srou
ji
54.0
54.0
64.8
64.8
73.2
85.5
85.5
Cur
rent
A
djus
ted
53.6
53.4
52.3
54.4
80.2
81.2
81.5
Klin
e
62.0
62.0
67.4
67.4
68.2
83.0
83.0
Srou
ji
54.3
54.3
62.0
62.0
68.8
81.6
81.6
Cur
rent
A
djus
ted
51.5
50.0
56.2
60.4
65.1
73.0
73.0
Test 1 2 1 2 1 1 2
Con
nect
ion
F1-3
/4-3
/8-1
6
F1-3
/4-3
/8-1
6
F1-5
/8-3
/8-1
6
F1-5
/8-3
/8-1
6
F2-3
/4-3
/8-1
6
F2-5
/8-3
/8-1
6
F2-5
/8-3
/8-1
6
Max
imum
Mom
ent (
ft-ki
ps)
Pred
icte
d St
reng
ths
App
lied
Mpr
ed/M
app
Page 69
61
The moment strengths adjusted by a factor of 1.2 correlated much better
with past results as Table 4.6 shows. Tables 4.5 and 4.6 show the average and standard
deviation of Mpred/Mapp for all three studies. The averages of Mpred/Mapp for Srouji et al.
and Kline et al. are 0.97 and 1.03, respectively, and the standard deviations are 0.025 and
0.105, respectively. Prior to adjustment, the average and standard deviation of Mpred/Mapp
for the current study were 1.15 and 0.117, respectively. After adjustment the average and
standard deviation of Mpred/Mapp for the current study are 0.96 and 0.097, respectively.
The average of 0.96 for adjusted values of Mpred/Mapp for the current study correlates
very well with averages of 0.97 for Srouji et al. and 1.03 for Kline et al.. The standard
deviation of 0.097 for the adjusted values of Mpred/Mapp for the current study falls within
the range of standard deviations for the previous studies (0.096 (Srouji) > 0.097 > 1.03
(Kline)). This shows the scatter of the adjusted data to be in an acceptable range.
Unfortunately, the load cell was damaged beyond repair during an experiment following
the completion of this study making a re-calibration impossible.
The adjustment of the moments for the current study had no affect on the total and
inelastic rotation capabilities of the specimens. This occurs because the rotations remain
unchanged even though the moments are increased.
Page 70
62
CHAPTER V
SUMMARY AND CONCLUSIONS
5.1 SUMMARY
An experimental investigation was conducted to evaluate the inelastic rotation
capability of flush end-plate moment connections under seismic loading. The inelastic
rotation of fully restrained connections in a steel moment frame during an earthquake is
used to dissipate the energy added to the structural system by earthquake ground motions.
Current design recommendations require steel moment connections in a seismic resisting
moment frame to resist a predetermined amount of inelastic rotation depending on the
classification of the moment frame. The three classifications of steel moment frames are
special, intermediate, and ordinary as discussed in Chapter I. The inelastic rotation
capability of the connection determines in which of the three types of steel moment
frames that the connection may be used.
Seven specimens representing four different flush end-plate configurations were
tested. The specimens were tested using slow cyclic or “quasi-static” loading. Two
different loading protocols were used in the study. Three tests were performed using the
loading protocol given in ATC-24 (1992), and four tests were performed using the
loading protocol given by SAC (1997). The connections were loaded using one of these
two loading protocols until additional load could not be applied.
The specimens were designed using yield-line theory to predict end-plate strength
and maximum bolt force predictions including prying action. The methods used to
Page 71
63
design the connections are presented in Chapter II. All connections were designed so that
the test beam was stronger than the connection elements, forcing the failure mechanism
to form in the connection elements. This proved successful as no failures occurred in the
test beams. End-plate fracture, weld fracture, or bolt fracture were observed at the end of
the tests.
Initial results for strength and stiffness did not correlate well with analytical
predictions or previous research. A possible explanation for this is a load cell calibration
problem. Using an adjustment factor of 1.2 the maximum moments applied to the
specimens were recomputed. The average of the ratios of the adjusted Mpred/Mapp for this
study is 0.96 with a standard deviation of 0.097. These adjusted results correlated well
with analytical predictions and previous research.
5.2 CONCLUSIONS
A summary of the inelastic rotation capabilities of the connections tested in this
study are presented in Table 5.1 along with the type of steel moment frame in which they
may be used. All of the connections with the exception of F1-5/8-3/8-16 test 1 exhibited
inelastic rotations greater than the 0.01 radians which is necessary for use in ordinary
steel moment frames. The two-bolt and four-bolt configurations were designed with one
“tight” and one “wide” configuration each. A tight configuration is one with small gages
and pitches while a wide configuration is one with larger gages and pitches. The tight
configurations were F1-5/8-3/8-16 and F2-5/8-3/8-16 and the wide configurations were
F1-3/4-3/8-16 and F2-3/4-3/8-16. It was found that the connections with a wide
Page 72
64
connection allowed for more yielding of the end-plate during testing. These connections
exhibited nearly double the inelastic rotation capability than their tight counterparts.
The inelastic rotation of these connections was achieved through the yielding of
the connection elements while the beam remained undamaged. The AISC seismic
specification (Seismic 1997) mandates that fully restrained connections in special and
intermediate moment frames be designed to have at least the same moment strength at the
column face as the plastic moment strength of the beam.
Connection TestInelasticRotation(radians)
ConnectionConfiguration(tight or wide)
*Type of Moment Framefor Which Connection isQualified as Defined byAISC (Seismic 1997)
F1-3/4-3/8-16 1 0.038 Wide Special (θi > 0.03)
F1-3/4-3/8-16 2 0.041 Wide Special (θi > 0.03)
F1-5/8-3/8-16 1 0.0093 Tight None
F1-5/8-3/8-16 2 0.013 Tight Ordinary (θi > 0.01)
F2-3/4-3/8-16 1 0.048 Wide Special (θi > 0.03)
F2-5/8-3/8-16 1 0.020 Tight Intermediate (θi > 0.02)
F2-5/8-3/8-16 2 0.018 Tight Ordinary (θi > 0.01)
*Inelastic Rotation Requirements are Shown in Parentheses
TABLE 5.1 Compliance With Inelastic Rotation Requirements
Page 73
65
Partially restrained connections, however, may achieve the required inelastic
rotation through the yielding of the connection elements. This requirement was written
with fully welded connections in mind and is conservative for bolted connections.
However, no exception for bolted connections is given in the specification. Therefore, by
the definitions outlined in the AISC seismic specification, the connections tested can only
be used as partially restrained connections in special and intermediate moment frames in
seismic areas. However, the tested connections are permitted for use as fully restrained
connections in ordinary moment frames because in ordinary moment frames the
connection shall be designed for a flexural strength equal to 1.1RyMp of the beam or the
maximum moment delivered to the system, whichever is less.
Two (F1-3/4-3/8-16 and F2-3/4-3/8) of the four connections tested in this study
satisfy the requirements for use as fully restrained connections in ordinary moment
frames in seismic areas. Because the AISC seismic specification (Seismic 1997) requires
that two full scale tests be done on the same connection configuration, F1-3/4-3/8-16
cannot be approved immediately for use in ordinary moment frames. The test performed
in this study shows, however, that this connection exhibits more than enough inelastic
rotation capability to be used in ordinary moment frames. Two tests were performed on
F1-5/8-16 with test 1 exhibiting enough inelastic rotation to be used in ordinary moment
frames while F1-5/8-3/8-16 test 2 did not. At this time the results for F1-5/8-3/8-16 are
inconclusive. Therefore it is recommended that further study be performed on this
connection configuration.
Page 74
66
Obviously due to cost and time constraints not every connection size and
configuration can be tested under seismic loading. Therefore AISC allows interpolation
and extrapolation from test data. AISC allows results of qualifying test to be extrapolated
for beam depths up to 10% greater than the tested beam depths. The beam depths in this
study were rather shallow, allowing the results of this study to be extrapolated for beam
depths of up to 16/0.9 or approximately 18 in.
Page 75
67
REFERENCES
Borgsmiller, J.T., and Murray, T.M., (1995) “Simplified Method For Design of MomentEnd-Plate Connections” Report No. CE/VPI-ST 95/19, Virginia Polytechnic Institute andState University, Blacksburg, Virginia.
Federal Emergency Management Agency, (1995) “Interim Guidelines: Evaluation,Repair, Modification and Design of Steel Moment Frames” FEMA 267 (SAC 96-02),FEMA, Washington, DC.
Ghobarah, A., Korol, R.M., and Osman, A. (1992) “Cyclic Behavior of Extended End-Plate Joints.” Journal of structural Engineering, Vol. 188(5), 1333-1353.
Ghobarah, A., Osman, A., Korol, R.M. (1990) “Behavior of Extended End-PlateConnection Under Cyclic Loading.” Engineering Structures, Vol. 12(1), 15-27.
“Guidelines for Cyclic Seismic Testing of Components of Steel Structures (ATC-24)”(1992) Applied Technology Council, Redwood City, CA
Hendrick, D.M., Murray, T.M. and Kukreti, A.R., (1985) “Unification of Flush End-PlateDesign Procedures” Report No. FSEL/MBMA 8305, University of Oklahoma, Norman,Oklahoma.
Kennedy, N.A., Vinnakota, S. and Sherbourne, A.N. (1981) “The Split-Tee Analogy inBolted Splices and Beam-Column Connections”, Proceedings of the InternationalConference on Joints in Structural Steelwork, 2.138-2.157
Kline, D.P., and Murray, T.M., and Rojiani K. (1989) “Performance of Snug Tight Boltsin Moment End-Plate Connections.” Report No. CE/VPI-ST 89/04, Virginia PolytechnicInstitute and State University, Blacksburg, Virginia.
Korol, R.M., Ghobarah, A., and Osman, A. (1990) “Extended End-Plate ConnectionsUnder Cyclic Loading. Behavior and Design.” Journal of Constructional SteelResearch, Vol. 16(4), 253-280.
Load and Resistance Factor Design Specification for Structural Steel Buildings (1993),American Institute of Steel Construction, Chicago, Illinois.
Manual of Steel Construction. (Load & Resistance Factor Design) (1994), SecondEdition, Vol. I and II, American Institute of Steel Construction, Chicago, Illinois.
Page 76
68
Meng, R.L., (1996) “Design of Moment End-Plate Connections For Seismic Loading.”Doctoral Dissertation, Department of Civil Engineering, Virginia Polytechnic Instituteand State University, Blacksburg, Virginia.
Phillips, J., and Morris, L.J., (1981) “The Effect of Plate Thickness on Flush End-PlateConnections”, Proceedings of the International Conference on Joints in Steelwork, 6.77-6.92.
Popov, E.P., and Bertero, V.V. (1989) “Performance of large Seismic Steel MomentConnections Under Cyclic Loads.” Engineering Journal, 12(2), 51-60.
Seismic Provisions for Structural Steel Buildings (1997), American Institute of SteelConstruction, Chicago, Illinois.
Srouji, R., Murray, T.M. and Kukreti, A.R., (1983) “Yield-Line Analysis of End-PlateConnections With Bolt Force Predictions” Report No. FSEL/MBMA 8305, University ofOklahoma, Norman, Oklahoma.
Tsai, K.C., and Popov. E. (1990) “Cyclic Behavior of End-Plate Moment Connections.”Journal of Structural Engineering., Vol. 166(11), 2917-2930.
Zarghamee, M.S., and Ojdrovic, R.P., (1995) “Northridge Postscript: Lessons on SteelConnections.” Civil Engineering, Vol. 65(4), 68-71.
Zoetermeijer, P., (1981) “Semi-Rigid Bolted Beam-to-Beam Column Connections WithStiffened Column Flanges and Flush End-Plates”, Proceedings of the InternationalConference on Joints in Steelwork, 2.99-2.118.
Page 77
69
APPENDIX A
NOMENCLATURE
Page 78
70
NOMENCLATURE
a - distance from bolt line to location of prying action
AISC - American Institute of Steel Construction
ATC - Applied Technology Council
B - bolt force
B1 - outer bolt force
B2 - inner bolt force
bf - beam flange width
ds - elemental length of line n
E - Young’s Modulus of Elasticity
F - flange force per bolt
Ff - total flange force
Fpy - end-plate material yield stress
Fyb - yield stress of the bolt
g - end-plate bolt gage distance
h - beam depth
Ln - length of yield line n
M1 - plastic moment at first hinge line to form
M2 - plastic moment at second hinge line to form
M3 - plastic moment at third hinge to form
Mapp - applied maximum moment
Mb - moment strength of the bolt
mp - plastic moment capacity of plate per unit length equal to (Fpytp2)/4
Mpl - moment strength of the connection calculated using yield-line theory
Mpred - predicted moment strength
Page 79
71
Mq - moment strength of the connection calculated using maximum bolt
force predictions
MU - end-plate ultimate moment capacity at beam end
N - number of yield lines in a mechanism
nj - the number of cycles to be performed in load step j
pb - distance between centerline of upper and lower rows of bolts
Pext - distance from face of beam flange to end of end-plate in Kennedy
model
pf - distance from bolt centerline to near face of beam flange
pt - distance from bolt centerline to far face of beam flange
Q - prying force
Qmax - maximum prying force
s - distance from bolt centerline to outermost yield-line
SAC - joint venture involving: the Structural Engineers Association of
California (SEAOC), Applied Technology Council (ATC), and the
California Universities for Research in Earthquake Engineering
(CUREe)
t1 - thick plate limit
t11 - thin plate limit
tp - end-plate thickness
w’ - width of end-plate per bolt at bolt line minus bolt hole diameter
We - total external work done on the connection
Wi - total internal energy stored
δ - deformation, a generic quantity including strains, angles of shear
distortion, rotations, axial deformations and displacements
δy - yield deformation
θ - interstory drift angle
θe - elastic rotation
θi - inelastic rotation
Page 80
72
θj - interstory driftat load step j in SAC protocol
θj - the peak deformation in load step j
θn - relative normal rotation of yield-line n
θt - total rotation
Page 81
73
APPENDIX B
TWO-BOLT FLUSH RESULTS AND TEST DATA
Page 82
74
SummaryF1-3/4-3/8-16
Test 1
Predicted Capacities:Moment Strength Predicted by Yield Line Analysis… … … … 51.5 ft-kipsMoment Strength Predicted by Bolt Rupture… … … … … … … 78.0 ft-kips
Maximum Moment From Test… … … … … … … … … … … … ..44.7 ft-kips
Maximum Total Rotation… … … … … … … … … … … … … … ..0.041 radMaximum Inelastic Rotation… … … … … … … … … … … … … .0.038 rad
Number of Completed Cycles… … … … … … … … … … … … ...28
Failure mode… … … … … … … … … … … … … … … … … … … ..End-plate fracturesFractures that occurred during test:• Top flange-to-end-plate weld, 1-3/8 in. in length on right side of the web.• Bottom flange-to-end-plate weld 1-1/4 in. in length on left side of the web.End-Plate fractures that occurred during test:• At web-to-end-plate weld 3-3/8 in. in length beginning at the top flange.• At web-to-end-plate weld 2-1/4 in. in length beginning at the bottom flange.
Page 83
75
Load and Displacement HistoryF1-3/4-3/8-16 Test 1
Load Step # Maximum Load(kips) Peak Deformation, θ Number of Cycles, n
1 3 * 4
2 4 * 3
3 5 * 3
4 * 0.011 3
5 * 0.0165 3
6 * 0.022 2
7 * 0.0275 2
8 * 0.033 2
9 * 0.0385 2
10 * 0.044 2
11 * 0.0465 2
*Note: Loading switched from load to deflection control at the fourth load step.ATC Loading Protocol Used
Material Testing ResultsF1-3/4-3/8-16 Test 1
Location % Elongation Fy(ksi)
Fu(ksi)
End-Plate 23.44 53.61 83.94
Beam Flange 22.67 60.59 85.16
Beam Web 21.32 73.77 82.74
Column Flange 24.92 57.61 85.80
Column Web 12.70 72.08 84.16
Page 84
76
F1-3/4-3/8-16 TEST 1 CONNECTION DETAILS
Page 85
77
Moment at End-Plate vs. Total Rotation at End-PlateF1-3/4-3/8-16 Test 1
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.05 -0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04 0.05Rotation (radians)
Moment (ft-kips)
Moment At End-Plate vs. Inelastic Rotation at End-PlateF1-3/4-3/8 Test 1
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.05 -0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04 0.05Rotation (radians)
Moment (ft-kips)
Page 86
78
Measured Bolt Strain X E vs. Flange Force - Bolt 1F1-3/4-3/8-16 Test 1
0
20
40
60
80
100
120
-40 -30 -20 -10 0 10 20 30 40Flange Force (kips)
Measured Strain x E (kips)
cycles 11-16
cycles 1-10
cycles 17-19
cycles 19-28
Measured Bolt Strain X E vs. Flange Force - Bolt 2F1-3/4-3/8-16 Test 1
0
40
80
120
160
200
-40 -30 -20 -10 0 10 20 30 40Flange Force (kips)
Measured Strain x E (kips)
cycles 1-10
cycles 11-13
cycles 14-16
cycles 17-19
Page 87
79
Measured Bolt Strain X E vs. Flange Force - Bolt 3F1-3/4-3/8-16 Test 1
0
20
40
60
80
100
120
140
160
-40 -30 -20 -10 0 10 20 30 40Flange Force (kips)
Measured Strain x E (kips)
cycles 1-10
cycles 11-16
cycles 17-
cycles 20-21
cycles 22-23
cycles 24-
Measured Bolt Strain X E vs. Flange Force - Bolt 4F1-3/4-3/8-16 Test 1
0
10
20
30
40
50
60
70
80
-40 -30 -20 -10 0 10 20 30 40Flange Force (kips)
Measured Strain x E (kips)
cycles 1-8
cycles 8-11
cycles 12-17
cycles 18-28
Page 88
80
Moment At End-Plate vs. Total Rotation at End-PlateF1-3/4-3/8-16 Test 1
0
10
20
30
40
50
60
70
80
90
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035Rotation (radians)
Moment (ft-kips)
TestPredictionYield Line Moment=51.5 ft-kipsMoment Due to Bolt Rupture=78.0 ft-kips
Page 89
81
SummaryF1-3/4-3/8-16
Test 2
Predicted Capacities:Moment Strength Predicted by Yield Line Analysis… … … … 50.0 ft-kipsMoment Strength Predicted by Bolt Rupture… … … … … … … 78.3 ft-kips
Maximum Moment From Test… … … … … … … … … … … … ..44.5 ft-kips
Maximum Total Rotation… … … … … … … … … … … … … … ..0.045 radMaximum Inelastic Rotation… … … … … … … … … … … … … .0.041 rad
Number of Completed Cycles… … … … … … … … … … … … ...17
Failure mode… … … … … … … … … … … … … … … … … … … ..End-plate fracturesFractures that occurred during test:• Top flange-to-end-plate weld on both sides of the web.End-Plate fractures that occurred during test:• At web-to-end-plate weld 3-3/8 in. in length beginning at the top flange.• At web-to-end-plate weld 2-3/8 in. in length beginning at the bottom flange.
Page 90
82
Load and Displacement HistoryF1-3/4-3/8-16 Test 2
Load Step # Maximum Load(kips) Peak Deformation, θ Number of
Cycles, n
1 3 * 4
2 4 * 3
3 6 * 3
4 * 0.022 3
5 * 0.030 3
6 * 0.045 ¼
*Note: Loading switched from load to deflection control at the fourth load step.ATC Loading Protocol Used
Material Testing ResultsF1-3/4-3/8-16 Test 2
Location % Elongation Fy
(ksi)Fu
(ksi)
End-Plate 23.44 52.62 82.71
Beam Flange 22.67 60.59 85.16
Beam Web 21.32 73.77 82.74
Column Flange 24.92 57.61 85.80
Column Web 12.70 72.08 84.16
Listing of Observations Made During Test
1. Flange-to-end-plate weld fracture noticed at cycle 17. Test Terminated
Page 91
83
F1-3/4-3/8-16 TEST 2 CONNECTION DETAILS
Page 92
84
Moment vs. Total Rotation at End-PlateF1-3/4-3/8-16 Test 2
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.05 -0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04 0.05Rotation (radians)
Moment (ft-kips)
Moment At End-Plate vs. Inelastic Rotation at End-PlateF1-3/4-3/8 Test 2
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.05 -0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04 0.05Rotation (radians)
Moment (ft-kips)
Page 93
85
Moment At End-Plate vs. End-Plate Separation at Top FlangeF1-3/4-3/8-16 Test 2
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.10 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90End-Plate Separation (in)
Moment (ft-kips)
Moment At End-Plate vs. End-Plate Separation at Bottom FlangeF1-3/4-3/8-16 Test 2
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.10 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90End-Plate Separation (in)
Moment (ft-kips)
Page 94
86
Measured Bolt Strain X E vs. Flange Force - Bolt 1F1-3/4-3/8-16 Test 2
0
20
40
60
80
100
120
140
160
-40 -30 -20 -10 0 10 20 30 40Flange Force (kips)
Measured Strain X E (kips)
Measured Bolt Strain X E vs. Flange Force - Bolt 2F1-3/4-3/8-16 Test 2
0
10
20
30
40
50
60
70
-40 -30 -20 -10 0 10 20 30 40Flange Force (kips)
Measured Strain X E (kips)
Page 95
87
Measured Bolt Strain X E vs. Flange Force - Bolt 3F1-3/4-3/8-16 Test 2
0
20
40
60
80
100
120
140
160
180
200
-40 -30 -20 -10 0 10 20 30 40Flange Force (kips)
Measured Strain X E (kips)
Measured Bolt Strain X E vs. Flange Force - Bolt 4F1-3/4-3/8-16 Test 2
0
10
20
30
40
50
60
70
80
90
100
-40 -30 -20 -10 0 10 20 30 40Flange Force (kips)
Measured Strain X E (kips)
Page 96
88
Moment At End-Plate vs. Total Rotation at End-PlateF1-3/4-3/8-16 Test 2
0
10
20
30
40
50
60
70
80
90
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035Rotation (radians)
Moment (ft-kips)
TestPredictionYield Line Moment=50.0 ft-kipsMoment Due to Bolt Rupture=78.3 ft-kips
Page 97
89
SummaryF1-5/8-3/8-16
Test 1
Predicted Capacities:Moment Strength Predicted by Yield Line Analysis… … … … 56.2 ft-kipsMoment Strength Predicted by Bolt Rupture… … … … … … … 57.1 ft-kips
Maximum Moment From Test… … … … … … … … … … … … ..43.6 ft-kips
Maximum Total Rotation… … … … … … … … … … … … … … ..0.016 radMaximum Inelastic Rotation… … … … … … … … … … … … … .0.0093 rad
Number of Completed Cycles… … … … … … … … … … … … ...26
Failure mode… … … … … … … … … … … … … … … … … … … ..Bolt FractureTop left bolt fractured during cycle 27
Page 98
90
Displacement HistoryF1-5/8-3/8-16 Test 1
Load Step # Peak Deformation, θ Number of Cycles, n
1 0.00375 6
2 0.005 6
3 0.0075 6
4 0.01 4
5 0.015 2
6 0.02 2
7 0.03 1/4
Material Testing ResultsF1-5/8-3/8-16 Test 1
Location % Elongation Fy(ksi)
Fu(ksi)
End-Plate 23.44 53.79 84.65
Beam Flange 22.26 60.19 86.16
Beam Web 20.28 73.86 82.63
Column Flange 24.14 57.77 85.21
Column Web 14.68 72.35 84.67
Listing of Observations Made During Test
1. Yielding of end-plate around bolts noticed during cycle 23.
2. At maximum positive moment of cycle 27, the top left bolt (i.e. the top left bolt at thetension flange) fractured. Test was terminated.
Page 99
91
F1-5/8-3/8-16 TEST 1 CONNECTION DETAILS
Page 100
92
Moment vs. Total Rotation at End-PlateF1-5/8-3/8-16 Test 1
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.020 -0.015 -0.010 -0.005 0.000 0.005 0.010 0.015 0.020Rotation (radians)
Moment (ft-kips)
Moment At End-Plate vs. Inelastic Rotation at End-PlateF1-5/8-3/8 Test 1
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.020 -0.015 -0.010 -0.005 0.000 0.005 0.010 0.015 0.020Rotation (radians)
Moment (ft-kips)
Page 101
93
Moment At End-Plate vs. End-Plate Separation at Top FlangeF1-5/8-3/8-16 Test 1
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.10 -0.05 0.00 0.05 0.10 0.15 0.20 0.25 0.30End-Plate Separation (in)
Moment (ft-kips)
Moment At End-Plate vs. End-Plate Separation at Bottom FlangeF1-5/8-3/8-16 Test 1
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.10 -0.05 0.00 0.05 0.10 0.15 0.20 0.25 0.30End-Plate Separation (in)
Moment (ft-kips)
Page 102
94
Moment At End-Plate vs. Total Rotation at End-PlateF1-5/8-3/8-16 Test 1
0
10
20
30
40
50
60
70
0.000 0.005 0.010 0.015 0.020 0.025Rotation (radians)
Moment (ft-kips)
TestPredictionYield Line Moment=56.2 ft-kipsMoment Due to Bolt Rupture=57.1 ft-kips
Page 103
95
SummaryF1-5/8-3/8-16
Test 2
Predicted Capacities:Moment Strength Predicted by Yield Line Analysis… … … … 60.4 ft-kipsMoment Strength Predicted by Bolt Rupture… … … … … … … 58.0 ft-kips
Maximum Moment From Test… … … … … … … … … … … … ..45.3 ft-kips
Maximum Total Rotation… … … … … … … … … … … … … … ..0.018 radMaximum Inelastic Rotation… … … … … … … … … … … … … .0.013 rad
Number of Completed Cycles… … … … … … … … … … … … ...26
Failure mode… … … … … … … … … … … … … … … … … … … ..Bolt FractureTop two bolts fractured during cycle 27
Page 104
96
Displacement HistoryF1-5/8-3/8-16 Test 2
Load Step # Peak Deformation, θ Number of Cycles, n
1 0.00375 6
2 0.005 6
3 0.0075 6
4 0.01 4
5 0.015 2
6 0.02 2
7 0.03 1/4
Material Testing ResultsF1-5/8-3/8-16 Test 2
Location % Elongation Fy(ksi)
Fu(ksi)
End-Plate 22.66 53.50 83.94
Beam Flange 22.26 60.19 86.16
Beam Web 20.28 73.86 82.63
Column Flange 24.14 57.77 85.21
Column Web 14.68 72.35 84.67
Listing of Observations Made During Test
1. Yielding of end-plate around bolts noticed during cycle 22.
2. At maximum positive moment of cycle 27, the top two bolts fractured. Test wasterminated.
Page 105
97
F1-5/8-3/8-16 TEST 2 CONNECTION DETAILS
Page 106
98
Moment vs. Total Rotation at End-PlateF1-5/8-3/8-16 Test 2
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.025 -0.020 -0.015 -0.010 -0.005 0.000 0.005 0.010 0.015 0.020Rotation (radians)
Moment (ft-kips)
Moment At End-Plate vs. Inelastic Rotation at End-PlateF1-5/8-3/8 Test 2
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.020 -0.015 -0.010 -0.005 0.000 0.005 0.010 0.015 0.020Rotation (radians)
Moment (ft-kips)
Page 107
99
Moment At End-Plate vs. End-Plate Separation at Top FlangeF1-5/8-3/8-16 Test 2
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.10 0.00 0.10 0.20 0.30 0.40 0.50 0.60End-Plate Separation (in)
Moment (ft-kips)
Moment At End-Plate vs. End-Plate Separation at Bottom FlangeF1-5/8-3/8-16 Test 2
-50
-40
-30
-20
-10
0
10
20
30
40
50
-0.10 0.00 0.10 0.20 0.30 0.40 0.50 0.60End-Plate Separation (in)
Moment (ft-kips)
Page 108
100
Moment At End-Plate vs. Total Rotation at End-PlateF1-5/8-3/8-16 Test 2
0
10
20
30
40
50
60
70
0.000 0.005 0.010 0.015 0.020 0.025Rotation (radians)
Moment (ft-kips)
TestPredictionYield Line Moment=60.4 ft-kipsMoment Due to Bolt Rupture=58.0 ft-kips
Page 109
101
APPENDIX C
FOUR-BOLT FLUSH RESULTS AND TEST DATA
Page 110
102
SummaryF2-3/4-3/8-16
Predicted Capacities:Moment Strength Predicted by Yield Line Analysis… … … … 65.1 ft-kipsMoment Strength Predicted by Bolt Rupture… … … … … … … 104.6 ft-kips
Maximum Moment From Test… … … … … … … … … … … … ..66.8 ft-kips
Maximum Total Rotation… … … … … … … … … … … … … … ..0.052 radMaximum Inelastic Rotation… … … … … … … … … … … … … .0.048 rad
Number of Completed Cycles… … … … … … … … … … … … ...12
Failure mode… … … … … … … … … … … … … … … … … … … ..End-plate and weldfracturesWeld fractures that occurred during test:• Top flange-to-end-plate on right side of the web.• Bottom flange-to-end-plate on left side of web.• Bottom flange-to-end-plate on right side of web, 2-3/8 in. in length beginning at the
edge of the flange.• Web-to-end-plate 6 in. in length beginning at the top flange.End-Plate Fractures that occurred during test:• At web-to-end-plate 6-1/2 in. in length beginning at the bottom flange.
Page 111
103
Load and Displacement HistoryF2-3/4-3/8-16
Load Step # Maximum Load(kips)
Peak Deformation,θ Number of Cycles, n
1 4.6 * 4
2 7.5 * 3
3 9.4 * 3
4 * 0.05 2
*Note: Loading switched from load to deflection control at the fourth load step.ATC Loading Protocol Used
Material Testing ResultsF2-3/4-3/8-16
Location % Elongation Fy(ksi)
Fu(ksi)
End-Plate 18.75 52.73 82.71
Beam Flange 23.95 60.94 86.94
Beam Web 21.09 74.39 83.28
Column Flange 24.41 58.35 86.18
Column Web 16.25 69.62 82.44
Listing of Observations Made During Test
1. Separation of end-plate noticed at top and bottom bolts during cycle 10.
2. End-plate-to-beam-web weld fracture noticed during cycle 10.
3. Test terminated after cycle 12
Page 112
104
F2-3/4-3/8-16 CONNECTION DETAILS
Page 113
105
Moment vs. Total Rotation at End-PlateF2-3/4-3/8-16
-80
-60
-40
-20
0
20
40
60
80
-0.06 -0.04 -0.02 0.00 0.02 0.04 0.06Rotation (radians)
Moment (ft-kips)
Moment At End-Plate vs. Inelastic Rotation at End-PlateF2-3/4-3/8-16
-80
-60
-40
-20
0
20
40
60
80
-0.06 -0.04 -0.02 0.00 0.02 0.04 0.06Rotation (radians)
Moment (ft-kips)
Page 114
106
Moment At End-Plate vs. End-Plate Separation at Top FlangeF2-3/4-3/8-16
-80
-60
-40
-20
0
20
40
60
80
-0.20 0.00 0.20 0.40 0.60 0.80 1.00End-Plate Separation (in)
Moment (ft-kips)
Moment At End-Plate vs. End-Plate Separation at Bottom FlangeF2-3/4-3/8-16
-80
-60
-40
-20
0
20
40
60
80
-0.20 0.00 0.20 0.40 0.60 0.80 1.00End-Plate Separation (in)
Moment (ft-kips)
Page 115
107
Measured Bolt Strain X E vs. Flange Force - Bolt 1F2-3/4-3/8-16
0
50
100
150
200
250
-60 -40 -20 0 20 40 60Flange Force (kips)
Measured Strain X E (kips)
Measured Bolt Strain X E vs. Flange Force - Bolt 2F2-3/4-3/8-16
0
10
20
30
40
50
60
70
-60 -40 -20 0 20 40 60Flange Force (kips)
Measured Strain X E (kips)
Page 116
108
Measured Bolt Strain X E vs. Flange Force - Bolt 3F2-3/4-3/8-16
0
5
10
15
20
25
30
35
40
-60 -40 -20 0 20 40 60
Flange Force (kips)
Measured Strain X E (kips)
Measured Bolt Strain X E vs. Flange Force - Bolt 4F2-3/4-3/8-16
0
5
10
15
20
25
30
35
40
45
-60 -40 -20 0 20 40 60Flange Force (kips)
Measured Strain X E (kips)
Page 117
109
Measured Bolt Strain X E vs. Flange Force - Bolt 5F2-3/4-3/8-16
0
5
10
15
20
25
30
35
40
45
-60 -40 -20 0 20 40 60Flange Force (kips)
Measured Strain X E (kips)
Measured Bolt Strain X E vs. Flange Force - Bolt 6F2-3/4-3/8-16
0
10
20
30
40
50
60
-60 -40 -20 0 20 40 60Flange Force (kips)
Measured Strain X E (kips)
Page 118
110
Measured Bolt Strain X E vs. Flange Force - Bolt 7F2-3/4-3/8-16
0
20
40
60
80
100
120
140
160
180
200
-60 -40 -20 0 20 40 60Flange Force (kips)
Measured Strain X E (kips)
Measured Bolt Strain X E vs. Flange Force - Bolt 8F2-3/4-3/8-16
0
20
40
60
80
100
120
-60 -40 -20 0 20 40 60Flange Force (kips)
Measured Strain X E (kips)
Page 119
111
Moment At End-Plate vs. Total Rotation at End-PlateF2-3/4-3/8-16
0
20
40
60
80
100
120
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035Rotation (radians)
Moment (ft-kips)
TestPredictionYield Line Moment=65.1 ft-kipsMoment Due to Bolt Rupture=104.6 ft-kips
Page 120
112
SummaryF2-5/8-3/8-16
Test 1
Predicted Capacities:Moment Strength Predicted by Yield Line Analysis… … … … 73.0 ft-kipsMoment Strength Predicted by Bolt Rupture… … … … … … … 76.8 ft-kips
Maximum Moment From Test… … … … … … … … … … … … ..67.7 ft-kips
Maximum Total Rotation… … … … … … … … … … … … … … ..0.028 radMaximum Inelastic Rotation… … … … … … … … … … … … … .0.02 rad
Number of Completed Cycles… … … … … … … … … … … … ...26
Failure mode… … … … … … … … … … … … … … … … … … … ..Bolt FractureBolt fractured during cycle 27
Page 121
113
Displacement HistoryF2-5/8-3/8-16 Test 1
Load Step # Peak Deformation, θ Number of Cycles, n
1 0.00375 6
2 0.005 6
3 0.0075 6
4 0.01 4
5 0.015 2
7 0.03 2
8 0.04 ½
Material Testing ResultsF2-5/8-3/8-16 Test 1
Location % Elongation Fy(ksi)
Fu(ksi)
End-Plate 23.02 52.97 82.89
Beam Flange 23.14 61.07 85.93
Beam Web 20.34 73.88 83.17
Column Flange 23.39 58.47 86.29
Column Web 15.75 70.69 84.70
Listing of Observations Made During Test
1. Slight end-plate separation noticed at both flanges during cycle 25.
2. At maximum negative moment of cycle 27, bolt fractured. Test was terminated.
Page 122
114
F2-5/8-3/8-16 TEST 1 CONNECTION DETAILS
Page 123
115
Moment vs. Total Rotation at End-PlateF2-5/8-3/8-16 Test 1
-80
-60
-40
-20
0
20
40
60
80
-0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04Rotation (radians)
Moment (ft-kips)
Moment At End-Plate vs. Inelastic Rotation at End-PlateF2-5/8-3/8 Test 1
-80
-60
-40
-20
0
20
40
60
80
-0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04Rotation (radians)
Moment (ft-kips)
Page 124
116
Moment At End-Plate vs. End-Plate Separation at Top FlangeF2-5/8-3/8-16 Test 1
-80
-60
-40
-20
0
20
40
60
80
-0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70End-Plate Separation (in)
Moment (ft-kips)
Moment At End-Plate vs. End-Plate Separation at Bottom FlangeF2-5/8-3/8-16 Test 1
-80
-60
-40
-20
0
20
40
60
80
-0.20 -0.10 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70End-Plate Separation (in)
Moment (ft-kips)
Page 125
117
Moment At End-Plate vs. Total Rotation at End-PlateF2-5/8-3/8-16 Test 1
0
20
40
60
80
100
120
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035Rotation (radians)
Moment (ft-kips)
TestPredictionYield Line Moment=73.0 ft-kipsMoment Due to Bolt Rupture=76.80 ft-kips
Page 126
118
DatapackTest F2-5/8-3/8-16
Test 2
Research on the Inelastic Rotation Capability of End-Plate MomentConnections
Mark R. BoorseResearch Assistant
Thomas M. MurrayPrincipal Investigator
Page 127
119
SummaryF2-5/8-3/8-16
Test 2
Predicted Capacities:Moment Strength Predicted by Yield Line Analysis… … … … 75.0 ft-kipsMoment Strength Predicted by Bolt Rupture… … … … … … … 76.8 ft-kips
Maximum Moment From Test… … … … … … … … … … … … ..67.9 ft-kips
Maximum Total Rotation… … … … … … … … … … … … … … ..0.033 radMaximum Inelastic Rotation… … … … … … … … … … … … … .0.018 rad
Number of Completed Cycles… … … … … … … … … … … … ...28
Failure mode… … … … … … … … … … … … … … … … … … … ..Bolt FractureBolt fractured during cycle 29
Page 128
120
Displacement HistoryF2-5/8-3/8-16 Test 2
Load Step # Peak Deformation, θ Number of Cycles, n
1 0.00375 6
2 0.005 6
3 0.0075 6
4 0.01 4
5 0.015 2
6 0.02 2
7 0.03 2
8 0.04 ¼
Material Testing ResultsF2-5/8-3/8-16 Test 2
Location % Elongation Fy(ksi)
Fu(ksi)
End-Plate 23.02 52.97 82.89
Beam Flange 23.14 61.07 85.93
Beam Web 20.34 73.88 83.17
Column Flange 23.39 58.47 86.29
Column Web 15.75 70.69 84.70
Listing of Observations Made During Test
1. Yielding of end-plate around bolts noticed during cycle 24.
2. At maximum negative moment of cycle 29, bolt fractured. Test was terminated.
Page 129
121
F2-5/8-3/8-16 TEST 2 CONNECTION DETAILS
Page 130
122
Moment vs. Total Rotation at End-PlateF2-5/8-3/8-16 Test 2
-80
-60
-40
-20
0
20
40
60
80
-0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04Rotation (radians)
Moment (ft-kips)
Moment At End-Plate vs. Inelastic Rotation at End-PlateF2-5/8-3/8 Test 2
-80
-60
-40
-20
0
20
40
60
80
-0.04 -0.03 -0.02 -0.01 0.00 0.01 0.02 0.03 0.04Rotation (radians)
Moment (ft-kips)
Page 131
123
Moment At End-Plate vs. End-Plate Separation at Top FlangeF2-5/8-3/8-16 Test 2
-80
-60
-40
-20
0
20
40
60
80
-0.20 -0.10 0.00 0.10 0.20 0.30 0.40End-Plate Separation (in)
Moment (ft-kips)
Moment At End-Plate vs. End-Plate Separation at Bottom FlangeF2-5/8-3/8-16 Test 2
-80
-60
-40
-20
0
20
40
60
80
-0.20 -0.10 0.00 0.10 0.20 0.30 0.40End-Plate Separation (in)
Moment (ft-kips)
Page 132
124
Moment At End-Plate vs. Total Rotation at End-PlateF2-5/8-3/8-16 Test 2
0
20
40
60
80
100
120
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035Rotation (radians)
Moment (ft-kips)
TestPredictionYield Line Moment=75.0 ft-kipsMoment Due to Bolt Rupture=76.8 ft-kips
Page 133
125
VITA
Mark R. Boorse was born in Brookfield, Wisconsin on July 30, 1974. After
graduating from high school in 1992 he entered the civil engineering program at the
University of Wisconsin at Platteville. In 1997 he attained a Bachelor of Science in Civil
Engineering and enrolled in the graduate program in the Civil Engineering Department at
Virginia Polytechnic Institute and State University, Blacksburg, Virginia.